Matching entries: 0
settings...
AuthorTitleYearJournal/ProceedingsReftypeDOI/URL
Jaśkowski, W. and Krawiec, K. How many dimensions in co-optimization 2011 Proceedings of the 13th Annual Conference Companion on Genetic and Evolutionary Computation, pp. 829-830  inproceedings URL 
Abstract: Co-optimization test-based problems is a class of tasks approached
typically with coevolutionary algorithms. It was recently shown that
such problems exhibit underlying objectives that form internal problem
structure, which can be extracted and analyzed in order to drive
the search or design better algorithms. The number of underlying
objectives is the dimension of the problem, which is of great importance,
since it may be a predictor of problem's directlyculty. In this paper,
we estimate the number of dimensions for Tic Tac Toe and the Density
Classification Task.
BibTeX:
@inproceedings{DBLP:conf/gecco/JaskowskiK11,
  author = {Wojciech Jaśkowski and Krzysztof Krawiec},
  title = {How many dimensions in co-optimization},
  booktitle = {Proceedings of the 13th Annual Conference Companion on Genetic and Evolutionary Computation},
  publisher = {ACM},
  year = {2011},
  pages = {829-830},
  url = {http://doi.acm.org/10.1145/2001858.2002110}
}
Krawiec, K. and Szubert, M.G. Learning n-tuple networks for othello by coevolutionary gradient search 2011 Proceedings of the 13th Annual Conference on Genetic and Evolutionary Computation, pp. 355-362  inproceedings DOI URL 
Abstract: We propose Coevolutionary Gradient Search, a blueprint for a family
of iterative learning algorithms that combine elements of local search
and population-based search. The approach is applied to learning
Othello strategies represented as n-tuple networks, using different
search operators and modes of learning. We focus on the interplay
between the continuous, directed, gradient-based search in the space
of weights, and fitness-driven, combinatorial, coevolutionary search
in the space of entire n-tuple networks. In an extensive experiment,
we assess both the objective and relative performance of algorithms,
concluding that the hybridization of search techniques improves the
convergence. The best algorithms not only learn faster than constituent
methods alone, but also produce top ranked strategies in the online
Othello League.
BibTeX:
@inproceedings{DBLP:conf/gecco/KrawiecS11,
  author = {Krzysztof Krawiec and Marcin Grzegorz Szubert},
  title = {Learning n-tuple networks for othello by coevolutionary gradient search},
  booktitle = {Proceedings of the 13th Annual Conference on Genetic and Evolutionary Computation},
  publisher = {ACM},
  year = {2011},
  pages = {355--362},
  url = {http://doi.acm.org/10.1145/2001576.2001626},
  doi = {http://doi.org/10.1145/2001576.2001626}
}
Błądek, I. and Krawiec, K. Evolutionary Program Sketching 2017 Genetic Programming: 20th European Conference, EuroGP 2017, Amsterdam, The Netherlands, April 19-21, 2017, Proceedings, pp. 3-18  inbook DOI URL 
BibTeX:
@inbook{Bladek2017,
  author = {Błądek, Iwo and Krawiec, Krzysztof},
  title = {Evolutionary Program Sketching},
  booktitle = {Genetic Programming: 20th European Conference, EuroGP 2017, Amsterdam, The Netherlands, April 19-21, 2017, Proceedings},
  publisher = {Springer International Publishing},
  year = {2017},
  pages = {3--18},
  url = {http://dx.doi.org/10.1007/978-3-319-55696-3_1},
  doi = {http://doi.org/10.1007/978-3-319-55696-3_1}
}
Pawlak, T.P. and Krawiec, K. Synthesis of Mathematical Programming Constraints with Genetic Programming 2017 Genetic Programming: 20th European Conference, EuroGP 2017, Amsterdam, The Netherlands, April 19-21, 2017, Proceedings, pp. 178-193  inbook DOI URL 
BibTeX:
@inbook{Pawlak2017,
  author = {Pawlak, Tomasz P. and Krawiec, Krzysztof},
  title = {Synthesis of Mathematical Programming Constraints with Genetic Programming},
  booktitle = {Genetic Programming: 20th European Conference, EuroGP 2017, Amsterdam, The Netherlands, April 19-21, 2017, Proceedings},
  publisher = {Springer International Publishing},
  year = {2017},
  pages = {178--193},
  url = {http://dx.doi.org/10.1007/978-3-319-55696-3_12},
  doi = {http://doi.org/10.1007/978-3-319-55696-3_12}
}
Pawlak, T.P. and Krawiec, K. Competent Geometric Semantic Genetic Programming for Symbolic Regression and Boolean Function Synthesis 2017 Evolutionary Computation
Vol. early access 
article DOI  
BibTeX:
@article{Pawlak:2017:ECJ,
  author = {Pawlak, Tomasz P. and Krawiec, Krzysztof},
  title = {Competent Geometric Semantic Genetic Programming for Symbolic Regression and Boolean Function Synthesis},
  journal = {Evolutionary Computation},
  publisher = {MIT Press},
  year = {2017},
  volume = {early access},
  doi = {http://doi.org/10.1162/EVCO_a_00205}
}
Pawlak, T.P. and Krawiec, K. Automatic Synthesis of Constraints from Examples using Mixed Integer Linear Programming 2017 European Journal on Operational Research
Vol. in press 
article DOI  
BibTeX:
@article{Pawlak:2017:EJOR,
  author = {Pawlak, Tomasz P. and Krawiec, Krzysztof},
  title = {Automatic Synthesis of Constraints from Examples using Mixed Integer Linear Programming},
  journal = {European Journal on Operational Research},
  publisher = {Elsevier},
  year = {2017},
  volume = {in press},
  doi = {http://doi.org/10.1016/j.ejor.2017.02.034}
}
Swan, J., Krawiec, K. and Ghani, N. Polytypic Genetic Programming 2017 Applications of Evolutionary Computation: 20th European Conference, EvoApplications 2017, Amsterdam, The Netherlands, April 19-21, 2017, Proceedings, Part II, pp. 66-81  inbook DOI URL 
BibTeX:
@inbook{Swan2017,
  author = {Swan, Jerry and Krawiec, Krzysztof and Ghani, Neil},
  title = {Polytypic Genetic Programming},
  booktitle = {Applications of Evolutionary Computation: 20th European Conference, EvoApplications 2017, Amsterdam, The Netherlands, April 19-21, 2017, Proceedings, Part II},
  publisher = {Springer International Publishing},
  year = {2017},
  pages = {66--81},
  url = {http://dx.doi.org/10.1007/978-3-319-55792-2_5},
  doi = {http://doi.org/10.1007/978-3-319-55792-2_5}
}
Krawiec, K. Evolutionary Feature Selection and Construction 2016 Encyclopedia of Machine Learning and Data Mining, pp. 1-5  inbook DOI URL 
BibTeX:
@inbook{Krawiec2016EncMachLearning,
  author = {Krawiec, Krzysztof},
  title = {Evolutionary Feature Selection and Construction},
  booktitle = {Encyclopedia of Machine Learning and Data Mining},
  publisher = {Springer US},
  year = {2016},
  pages = {1--5},
  url = {http://dx.doi.org/10.1007/978-1-4899-7502-7_90-1},
  doi = {http://doi.org/10.1007/978-1-4899-7502-7_90-1}
}
Krawiec, K. Behavioral Program Synthesis with Genetic Programming 2016
Vol. 618 
book DOI URL 
BibTeX:
@book{KrawiecBPS2015,
  author = {Krzysztof Krawiec},
  title = {Behavioral Program Synthesis with Genetic Programming},
  publisher = {Springer International Publishing},
  year = {2016},
  volume = {618},
  note = { http://www.cs.put.poznan.pl/kkrawiec/bps },
  url = { http://www.springer.com/gp/book/9783319275635 },
  doi = {http://doi.org/10.1007/978-3-319-27565-9}
}
Krawiec, K. Behavioral Program Synthesis with Genetic Programming 2016
Vol. 618 
book  
BibTeX:
@book{KrawiecBPS2015short,
  author = {Krzysztof Krawiec},
  title = {Behavioral Program Synthesis with Genetic Programming},
  publisher = {Springer},
  year = {2016},
  volume = {618}
}
Krawiec, K. and Swan, J. Distance Metric Ensemble Learning and the Andrews-Curtis Conjecture 2016   online URL 
Abstract: Motivated by the search for a counterexample to the Poincaré conjecture in three and four dimensions, the Andrews-Curtis conjecture was proposed in 1965. It is now generally suspected that the Andrews-Curtis conjecture is false, but small potential counterexamples are not so numerous, and previous work has attempted to eliminate some via combinatorial search. Progress has however been limited, with the most successful approach (breadth-first-search using secondary storage) being neither scalable nor heuristically-informed. A previous empirical analysis of problem structure examined several heuristic measures of search progress and determined that none of them provided any useful guidance for search. In this article, we induce new quality measures directly from the problem structure and combine them to produce a more effective search driver via ensemble machine learning. By this means, we eliminate 19 potential counterexamples, the status of which had been unknown for some years.
BibTeX:
@online{KrawiecSwan:2016:AndrewsCurtis,
  author = {Krawiec, Krzysztof and Swan, Jerry},
  title = {Distance Metric Ensemble Learning and the Andrews-Curtis Conjecture},
  year = {2016},
  url = {http://arxiv.org/abs/1606.01412}
}
Krawiec, K., Swan, J. and O’Reilly, U.-M. Behavioral Program Synthesis: Insights and Prospects 2016 Genetic Programming Theory and Practice XIII  incollection DOI  
Abstract: Genetic programming (GP) is a stochastic, iterative generate-and-test approach to synthesizing programs from tests, i.e. examples of the desired input-output mapping. The number of passed tests, or the total error in continuous domains, is a natural objective measure of a program's performance and a common yardstick when experimentally comparing algorithms. In GP, it is also by default used to guide the evolutionary search process. An assumption that an objective function should also be an efficient `search driver' is common for all metaheuristics, such as the evolutionary algorithms which GP is a member of. Programs are complex combinatorial structures that exhibit even more complex input-output behavior, and in this chapter we discuss why this complexity cannot be effectively reflected by a single scalar objective. In consequence, GP algorithms are systemically `underinformed' about the characteristics of programs they operate on, and pay for this with unsatisfactory performance and limited scalability. This chapter advocates behavioral program synthesis, where programs are characterized by informative execution traces that enable multifaceted evaluation and substantially change the roles of components in an evolutionary infrastructure. We provide a unified perspective on past work in this area, discuss the consequences of the behavioral viewpoint, outlining the future avenues for program synthesis and the wider application areas that lie beyond.
BibTeX:
@incollection{Krawiec:2015:GPTP,
  author = {Krzysztof Krawiec and Jerry Swan and Una-May O’Reilly},
  title = {Behavioral Program Synthesis: Insights and Prospects},
  booktitle = {Genetic Programming Theory and Practice XIII},
  publisher = {Springer},
  year = {2016},
  doi = {http://doi.org/10.1007/978-3-319-34223-8_10}
}
Liskowski, P. and Krawiec, K. Surrogate Fitness via Factorization of Interaction Matrix 2016
Vol. 9594EuroGP 2016: Proceedings of the 19th European Conference on Genetic Programming, pp. 65-79 
inproceedings  
Abstract: We propose SFIMX, a method that reduces the number of
required interactions between programs and tests in
genetic programming. SFIMX performs factorization of
the matrix of the outcomes of interactions between the
programs in a working population and the tests.
Crucially, that factorization is applied to matrix that
is only partially filled with interaction outcomes,
i.e., sparse. The reconstructed approximate interaction
matrix is then used to calculate the fitness of
programs. In empirical comparison to several reference
methods in categorical domains, SFIMX attains higher
success rate of synthesizing correct programs within a
given computational budget.
BibTeX:
@inproceedings{Liskowski:2016:EuroGP,
  author = {Pawel Liskowski and Krzysztof Krawiec},
  title = {Surrogate Fitness via Factorization of Interaction Matrix},
  booktitle = {EuroGP 2016: Proceedings of the 19th European Conference on Genetic Programming},
  publisher = {Springer Verlag},
  year = {2016},
  volume = {9594},
  pages = {65--79}
}
Liskowski, P. and Krawiec, K. Segmenting Retinal Blood Vessels with Deep Neural Networks 2016 IEEE Transactions on Medical Imaging
Vol. PP(99), pp. 1-1 
article DOI  
BibTeX:
@article{Liskowski:2016:IEEETMI,
  author = {Paweł Liskowski and Krzysztof Krawiec},
  title = {Segmenting Retinal Blood Vessels with Deep Neural Networks},
  journal = {IEEE Transactions on Medical Imaging},
  year = {2016},
  volume = {PP},
  number = {99},
  pages = {1-1},
  doi = {http://doi.org/10.1109/TMI.2016.2546227}
}
Liskowski, P. and Krawiec, K. Online Discovery of Search Objectives for Test-based Problems 2016 Evolutionary Computation  article  
Abstract:
In test-based problems, commonly approached with competitive coevolutionary algorithms, the fitness of a candidate solution is determined by the outcomes of its interactions with multiple tests. Usually, fitness is a scalar aggregate of interaction outcomes, and as such imposes a complete order on the candidate solutions. However, passing different tests may require unrelated `skills', and candidate solutions may vary with respect to such capabilities. In this paper, we provide theoretical evidence that scalar fitness, inherently incapable of capturing such differences, is likely to lead to premature convergence. To mitigate this problem, we propose disco, a method that automatically identifies the groups of tests for which the candidate solutions behave similarly and define the above skills. Each such group gives rise to a derived objective, which together guide the search algorithm in multi-objective fashion. When applied to several well-known test-based problems, the proposed approach significantly outperforms the conventional two-population coevolution. This opens the door to efficient and generic countermeasures to premature convergence for both coevolutionary and evolutionary algorithms applied to problems featuring aggregating fitness functions.
BibTeX:
@article{Liskowski:2016:EC,
  author = {Paweł Liskowski and Krzysztof Krawiec},
  title = {Online Discovery of Search Objectives for Test-based Problems},
  journal = {Evolutionary Computation},
  year = {2016}
}
Pawlak, T.P. and Krawiec, K. Semantic Geometric Initialization 2016
Vol. 9594EuroGP 2016: Proceedings of the 19th European Conference on Genetic Programming, pp. 254-269 
inproceedings  
Abstract: A common approach in Geometric Semantic Genetic
Programming (GSGP) is to seed initial populations using
conventional, semantic-unaware methods like Ramped
Half-and-Half. We formally demonstrate that this may
limit GSGP's ability to find a program with the sought
semantics. To overcome this issue, we determine the
desired properties of geometric-aware semantic
initialization and implement them in Semantic Geometric
Initialization (Sgi) algorithm, which we
instantiate for symbolic regression and Boolean
function synthesis problems. Properties of Sgi
and its impact on GSGP search are verified
experimentally on nine symbolic regression and nine
Boolean function synthesis benchmarks. When assessed
experimentally, Sgi leads to superior
performance of GSGP search: better best-of-run fitness
and higher probability of finding the optimal
program.
BibTeX:
@inproceedings{Pawlak2:2016:EuroGP,
  author = {Tomasz P. Pawlak and Krzysztof Krawiec},
  title = {Semantic Geometric Initialization},
  booktitle = {EuroGP 2016: Proceedings of the 19th European Conference on Genetic Programming},
  publisher = {Springer Verlag},
  year = {2016},
  volume = {9594},
  pages = {254--269}
}
Swan, J. and Krawiec, K. Discovering Relational Structure in Program Synthesis Problems with Analogical Reasoning 2016 Genetic Programming Theory and Practice XIV  incollection  
Abstract: Much recent progress in Genetic Programming (GP) can be ascribed to work in semantic GP, which facilitates program induction by considering program behavior on individual fitness cases. It is therefore interesting to consider whether alternative decompositions of fitness cases might also provide useful information. The one we present here is motivated by work in analogical reasoning. So-called proportional analogies (`gills are to fish as lungs are to mammals') have a hierarchical relational structure that can be captured using the formalism of Structural Information Theory. We show how proportional analogy problems can be solved with GP and, conversely, how analogical reasoning can be engaged in GP to provide for problem decomposition. The idea is to treat pairs of fitness cases as if they formed a proportional analogy problem, identify relational consistency between them, and use it to inform the search process.
BibTeX:
@incollection{Swan:2016:GPTP,
  author = {Jerry Swan and Krzysztof Krawiec},
  title = {Discovering Relational Structure in Program Synthesis Problems with Analogical Reasoning},
  booktitle = {Genetic Programming Theory and Practice XIV},
  publisher = {Springer},
  year = {2016}
}
Szkulmowski, M., Rumiński, D., Liskowski, P., Wieloch, B., Krawiec, K., Sikorski, B. and Wojtkowski, M.D. OCT retinal angiography using neural networks 2016 The Annual Meeting of the Association for Research in Vision and Ophthalmology  inproceedings URL 
BibTeX:
@inproceedings{arvo:2016:oct,
  author = {Maciej Szkulmowski and Daniel Rumiński and Paweł Liskowski and Bartosz Wieloch and Krzysztof Krawiec and Bartosz Sikorski and Maciej D. Wojtkowski},
  title = {OCT retinal angiography using neural networks},
  booktitle = {The Annual Meeting of the Association for Research in Vision and Ophthalmology},
  year = {2016},
  url = { http://www.arvo.org/webs/am2016/sectionpdf/MOI/Session_130.pdf }
}
Wyciszkiewicz, A., Pawlak, M.A. and Krawiec, K. Cerebellar Volume in Children With Attention-Deficit Hyperactivity Disorder (ADHD): Replication Study 2016 J. Child Neurol.  article DOI  
Abstract: Attention Deficit Hyperactivity Disorder (ADHD) is associated with altered cerebellar volume and cerebellum is associated with cognitive performance. However there are mixed results regarding the cerebellar volume in young patients with ADHD. To clarify the size and direction of this effect, we conducted the analysis on the large public database of brain images. The aim of this study was to confirm that cerebellar volume in ADHD is smaller than in control subjects in currently the largest publicly available cohort of ADHD subjects.We applied cross-sectional case control study design by comparing 286 ADHD patients (61 female) with age and gender matched control subjects. Volumetric measurements of cerebellum were obtained using automated segmentation with FreeSurfer 5.1. Statistical analysis was performed in R-CRAN statistical environment. Patients with ADHD had significantly smaller total cerebellar volumes (134.5±17.11cm(3) vs.138.90±15.32 cm(3)). The effect was present in both females and males (males 136.9±14.37 cm(3) vs. 141.20±14.75 cm(3); females 125.7±12.34 cm(3) vs. 131.20±15.03 cm(3)). Age was positively and significantly associated with the cerebellar volumes. These results indicate either delayed or disrupted cerebellar development possibly contributing to ADHD pathophysiology.
BibTeX:
@article{Krawiec2016JChildNeurology,
  author = {Wyciszkiewicz, A. and Pawlak, M. A. and Krawiec, K. },
  title = {Cerebellar Volume in Children With Attention-Deficit Hyperactivity Disorder (ADHD): Replication Study},
  journal = {J. Child Neurol.},
  year = {2016},
  note = { PubMed:https://www.ncbi.nlm.nih.gov/pubmed/27888270 },
  doi = {http://doi.org/10.1177/0883073816678550}
}
Fitzgerald, J., Ryan, C., Medernach, D. and Krawiec, K. An Integrated Approach to Stage 1 Breast Cancer Detection 2015 Proceedings of the 2015 Annual Conference on Genetic and Evolutionary Computation, pp. 1199-1206  inproceedings DOI URL 
Abstract: We present an automated, end-to-end approach for Stage 1 breast cancer detection. The first phase of our proposed workflow takes individual digital mammograms as input and outputs several smaller subimages from which the background has been removed. Next, we extract a set of features which capture textural information from the segmented images. In the final phase, the most salient of these features are fed into a Multi-Objective Genetic Programming system which then evolves classifiers capable of identifying those segments which may have suspicious areas that require further investigation.
A key aspect of this work is the examination of several new experimental configurations which focus on textural asymmetry between breasts. The best evolved classifier using such a configuration can deliver results of 100 percent accuracy on true positives and a false positive per image rating of just 0.33, which is better than the current state of the art.
BibTeX:
@inproceedings{Fitzgerald2015Gecco,
  author = {Fitzgerald, Jeannie and Ryan, Conor and Medernach, David and Krawiec, Krzysztof},
  title = {An Integrated Approach to Stage 1 Breast Cancer Detection},
  booktitle = {Proceedings of the 2015 Annual Conference on Genetic and Evolutionary Computation},
  publisher = {ACM},
  year = {2015},
  pages = {1199--1206},
  url = {http://doi.acm.org/10.1145/2739480.2754761},
  doi = {http://doi.org/10.1145/2739480.2754761}
}
Jakowski, W., Szubert, M., Liskowski, P. and Krawiec, K. High-Dimensional Function Approximation for Knowledge-Free Reinforcement Learning: a Case Study in SZ-Tetris 2015 Proceedings of the 2015 Annual Conference on Genetic and Evolutionary Computation, pp. 567-573  inproceedings DOI URL 
Abstract: SZ-Tetris, a restricted version of
Tetris, is a difficult reinforcement learning task. Previous research
showed that, similarly to the original Tetris, value function-based
methods such as temporal difference learning, do not work well for
SZ-Tetris. The best performance in this game was achieved by employing
direct policy search techniques, in particular the cross-entropy
method in combination with handcrafted features. Nonetheless, a simple
heuristic hand-coded player scores even higher. Here we show that it
is possible to equal its performance with CMA-ES (Covariance Matrix
Adaptation Evolution Strategy). We demonstrate that further
improvement is possible by employing systematic n-tuple network, a
knowledge-free function approximator, and VD-CMA-ES, a linear variant
of CMA-ES for high dimension optimization. Last but not least, we show
that a large systematic n-tuple network (involving more than 4 million
parameters) allows the classical temporal difference learning
algorithm to obtain similar average performance to VD-CMA-ES, but at
20 times lower computational expense, leading to the best policy for
SZ-Tetris known to date. These results enrich the current
understanding of difficulty of SZ-Tetris, and shed new light on the
capabilities of particular search paradigms when applied to
representations of various characteristics and dimensionality.
BibTeX:
@inproceedings{Jaskowski2015sztetris,
  author = {Wojciech Jakowski and Marcin Szubert and Paweł Liskowski and Krzysztof Krawiec},
  title = {High-Dimensional Function Approximation for Knowledge-Free Reinforcement Learning: a Case Study in SZ-Tetris},
  booktitle = {Proceedings of the 2015 Annual Conference on Genetic and Evolutionary Computation},
  publisher = {ACM},
  year = {2015},
  pages = {567--573},
  url = {http://doi.acm.org/10.1145/2739480.2754783},
  doi = {http://doi.org/10.1145/2739480.2754783}
}
Krawiec, K. FUEL = FUnctional Evolutionary aLgorithms 2015   online URL 
BibTeX:
@online{KrawiecFUEL2016,
  author = {Krzysztof Krawiec},
  title = {FUEL = FUnctional Evolutionary aLgorithms},
  year = {2015},
  note = {https://github.com/kkrawiec/fuel},
  url = {https://github.com/kkrawiec/fuel}
}
Krawiec, K. and Liskowski, P. Automatic Derivation of Search Objectives for Test-Based Genetic Programming 2015
Vol. 902518th European Conference on Genetic Programming, pp. 53-65 
inproceedings DOI  
Abstract: In genetic programming (GP), programs are usually
evaluated by applying them to tests, and fitness
function indicates only how many of them have been
passed. We posit that scrutinising the outcomes of
programs interactions with individual tests may help
making program synthesis more effective. To this aim,
we propose DOC, a method that autonomously derives new
search objectives by clustering the outcomes of
interactions between programs in the population and the
tests. The derived objectives are subsequently used to
drive the selection process in a single or
multiobjective fashion. An extensive experimental
assessment on 15 discrete program synthesis tasks
representing two domains shows that DOC significantly
outperforms conventional GP and implicit fitness
sharing.
BibTeX:
@inproceedings{Krawiec:2015:EuroGP,
  author = {Krzysztof Krawiec and Pawel Liskowski},
  title = {Automatic Derivation of Search Objectives for Test-Based Genetic Programming},
  booktitle = {18th European Conference on Genetic Programming},
  publisher = {Springer},
  year = {2015},
  volume = {9025},
  pages = {53--65},
  doi = {http://doi.org/10.1007/978-3-319-16501-1_5}
}
Krawiec, K. and Pawlak, M. Genetic Programming with Alternative Search Drivers for Detection of Retinal Blood Vessels 2015
Vol. 902818th European Conference on the Applications of Evolutionary Computation, pp. 554-566 
inproceedings DOI  
Abstract: A classification task is a test-based problem, with
examples corresponding to tests. A correct
classification is equivalent to passing a test, while
incorrect to failing it. This applies also to
classifying pixels in an image, viz. image
segmentation. A natural performance indicator in such a
setting is the accuracy of classification, i.e., the
fraction of passed tests. When solving a classification
tasks with genetic programming, it is thus common to
employ this indicator as a fitness function. However,
recent developments in GP as well as some earlier work
suggest that the quality of evolved solutions may
benefit from using other search drivers to guide the
traversal of the space of programs. In this study, we
systematically verify the usefulness of selected
alternative search drivers in the problem of detection
of blood vessels in ophthalmology imaging.
BibTeX:
@inproceedings{Krawiec:2015:evoApplications,
  author = {Krzysztof Krawiec and Mikołaj Pawlak},
  title = {Genetic Programming with Alternative Search Drivers for Detection of Retinal Blood Vessels},
  booktitle = {18th European Conference on the Applications of Evolutionary Computation},
  publisher = {Springer},
  year = {2015},
  volume = {9028},
  pages = {554--566},
  doi = {http://doi.org/10.1007/978-3-319-16549-3_45}
}
Liskowski, P., Krawiec, K., Helmuth, T. and Spector, L. Comparison of Semantic-aware Selection Methods in Genetic Programming 2015 Proceedings of the Companion Publication of the 2015 on Genetic and Evolutionary Computation Conference, pp. 1301-1307  inproceedings DOI URL 
Abstract: This study investigates the performance of several semantic-aware selection methods for genetic programming (GP). In particular, we consider methods that do not rely on complete GP semantics (i.e., a tuple of outputs produced by a program for fitness cases (tests)), but on binary outcome vectors that only state whether a given test has been passed by a program or not. This allows us to relate to test-based problems commonly considered in the domain of coevolutionary algorithms and, in prospect, to address a wider range of practical problems, in particular the problems where desired program output is unknown (e.g., evolving GP controllers). The selection methods considered in the paper include implicit fitness sharing (ifs), discovery of derived objectives (doc), lexicase selection (lex), as well as a hybrid of the latter two. These techniques, together with a few variants, are experimentally compared to each other and to conventional GP on a battery of discrete benchmark problems. The outcomes indicate superior performance of lex and ifs, with some variants of doc showing certain potential.
BibTeX:
@inproceedings{Liskowski:GECCO:2015,
  author = {Liskowski, Paweł and Krawiec, Krzysztof and Helmuth, Thomas and Spector, Lee},
  title = {Comparison of Semantic-aware Selection Methods in Genetic Programming},
  booktitle = {Proceedings of the Companion Publication of the 2015 on Genetic and Evolutionary Computation Conference},
  publisher = {ACM},
  year = {2015},
  pages = {1301--1307},
  url = {http://doi.acm.org/10.1145/2739482.2768505},
  doi = {http://doi.org/10.1145/2739482.2768505}
}
Pawlak, T.P. and Krawiec, K. Progress properties and fitness bounds for geometric semantic search operators 2015 Genetic Programming and Evolvable Machines, pp. 1-19  article DOI URL 
Abstract:
Metrics are essential for geometric semantic genetic programming. On one hand, they structure the semantic space and govern the behavior of geometric search operators; on the other, they determine how fitness is calculated. The interactions between these two types of metrics are an important aspect that to date was largely neglected. In this paper, we investigate these interactions and analyze their consequences. We provide a systematic theoretical analysis of the properties of abstract geometric semantic search operators under Minkowski metric of an arbitrary order. For nine combinations of popular metrics (city-block, Euclidean, and Chebyshev) used in fitness function and by search operators, we derive pessimistic bounds on fitness change. We also define three types of progress properties (weak, potential, and strong) and verify them for operators under those metrics. The analysis allows us to determine the combinations of metrics that are most attractive in terms of progress properties and deterioration bounds.
BibTeX:
@article{Pawlak:2015:GPEM,
  author = {Pawlak, Tomasz P. and Krawiec, Krzysztof},
  title = {Progress properties and fitness bounds for geometric semantic search operators},
  journal = {Genetic Programming and Evolvable Machines},
  publisher = {Springer US},
  year = {2015},
  pages = {1-19},
  url = {http://dx.doi.org/10.1007/s10710-015-9252-6},
  doi = {http://doi.org/10.1007/s10710-015-9252-6}
}
Pawlak, T.P., Wieloch, B. and Krawiec, K. Semantic Backpropagation for Designing Search Operators in Genetic Programming 2015 IEEE Transactions on Evolutionary Computation
Vol. 19(3), pp. 326-340 
article DOI URL 
Abstract: In genetic programming, a search algorithm is expected to produce a program that achieves the desired final computation state (desired output). To reach that state, an executing program needs to traverse certain intermediate computation states. An evolutionary search process is expected to autonomously discover such states. This can be difficult for nontrivial tasks that require long programs to be solved. The semantic backpropagation algorithm proposed in this paper heuristically inverts the execution of evolving programs to determine the desired intermediate computation states. Two search operators, Random Desired Operator and Approximately Geometric Semantic Crossover, use the intermediate states determined by semantic backpropagation to define subtasks of the original programming task, which are then solved using an exhaustive search. The operators outperform the standard genetic search operators and other semantic-aware operators when compared on a suite of symbolic regression and Boolean benchmarks. This result and additional analysis conducted in this study indicate that semantic backpropagation helps evolution at identifying the desired intermediate computation states, and makes the search process more efficient.
BibTeX:
@article{KrawiecPawlakWielochTEVC14,
  author = {Pawlak, Tomasz P. and Wieloch, Bartosz and Krawiec, Krzysztof},
  title = {Semantic Backpropagation for Designing Search Operators in Genetic Programming},
  journal = {IEEE Transactions on Evolutionary Computation},
  year = {2015},
  volume = {19},
  number = {3},
  pages = {326-340},
  url = { http://www.cs.put.poznan.pl/kkrawiec/pubs/TEVC2321259.pdf },
  doi = {http://doi.org/10.1109/TEVC.2014.2321259}
}
Pawlak, T.P., Wieloch, B. and Krawiec, K. Review and comparative analysis of geometric semantic crossovers 2015 Genetic Programming and Evolvable Machines
Vol. 16(3), pp. 351-386 
article DOI URL 
Abstract: This paper provides a structured, unified, formal and empirical perspective on all geometric semantic crossover operators proposed so far, including the exact geometric crossover by Moraglio, Krawiec, and Johnson, as well as the approximately geometric operators. We start with presenting the theory of geometric semantic genetic programming, and discuss the implications of geometric operators for the structure of fitness landscape. We prove that geometric semantic crossover can by construction produce an offspring that is not worse than the fitter parent, and that under certain conditions such an offspring is guaranteed to be not worse than the worse parent. We review all geometric semantic crossover operators presented to date in the literature, and conduct a comprehensive experimental comparison using a tree-based genetic programming framework and a representative suite of nine symbolic regression and nine Boolean function synthesis tasks. We scrutinize the performance (program error and success rate), generalization, computational cost, bloat, population diversity, and the operators' capability to generate geometric offspring. The experiment leads to several interesting conclusions, the primary one being that an operator's capability to produce geometric offspring is positively correlated with performance. The paper is concluded by recommendations regarding the suitability of operators for the particular domains of program induction tasks.
BibTeX:
@article{PawlakWielochKrawiec:2015:GPEM,
  author = {Pawlak, Tomasz P. and Wieloch, Bartosz and Krawiec, Krzysztof},
  title = {Review and comparative analysis of geometric semantic crossovers},
  journal = {Genetic Programming and Evolvable Machines},
  year = {2015},
  volume = {16},
  number = {3},
  pages = {351--386},
  url = {http://dx.doi.org/10.1007/s10710-014-9239-8},
  doi = {http://doi.org/10.1007/s10710-014-9239-8}
}
Rumiński, D., Sikorski, B.L., Bukowska, D., Szkulmowski, M., Krawiec, K., Malukiewicz, G., Bieganowski, L. and Wojtkowski, M. OCT angiography by absolute intensity difference applied to normal and diseased human retinas 2015 Biomedical Optics Express
Vol. 6(8), pp. 2738-2754 
article DOI URL 
Abstract: We compare four optical coherence tomography techniques for noninvasive visualization of microcapillary network in the human retina and murine cortex. We perform phantom studies to investigate contrast-to-noise ratio for angiographic images obtained with each of the algorithm. We show that the computationally simplest absolute intensity difference angiographic OCT algorithm that bases only on two cross-sectional intensity images may be successfully used in clinical study of healthy eyes and eyes with diabetic maculopathy and branch retinal vein occlusion.
BibTeX:
@article{Ruminski:15,
  author = {Daniel Rumiński and Bartosz L. Sikorski and Danuta Bukowska and Maciej Szkulmowski and Krzysztof Krawiec and Grażyna Malukiewicz and Lech Bieganowski and Maciej Wojtkowski},
  title = {OCT angiography by absolute intensity difference applied to normal and diseased human retinas},
  journal = {Biomedical Optics Express},
  publisher = {OSA},
  year = {2015},
  volume = {6},
  number = {8},
  pages = {2738--2754},
  url = {http://www.osapublishing.org/boe/abstract.cfm?URI=boe-6-8-2738},
  doi = {http://doi.org/10.1364/BOE.6.002738}
}
Ryan, C., Fitzgerald, J., Krawiec, K. and Medernach, D. Handbook of Genetic Programming Applications 2015 , pp. 245-287  inbook DOI URL 
BibTeX:
@inbook{Fitzgerald2015,
  author = {Ryan, Conor and Fitzgerald, Jeannie and Krawiec, Krzysztof and Medernach, David},
  title = {Handbook of Genetic Programming Applications},
  publisher = {Springer International Publishing},
  year = {2015},
  pages = {245--287},
  url = {http://dx.doi.org/10.1007/978-3-319-20883-1_10},
  doi = {http://doi.org/10.1007/978-3-319-20883-1_10}
}
Stanislawska, K., Krawiec, K. and Vihma, T. Genetic Programming for Estimation of Heat Flux Between the Atmosphere and Sea Ice in Polar Regions 2015 Proceedings of the 2015 Annual Conference on Genetic and Evolutionary Computation, pp. 1279-1286  inproceedings DOI URL 
Abstract: The Earth surface and atmosphere exchange heat via turbulent fluxes. An accurate description of the heat exchange is essential in modelling the weather and climate. In these models the heat fluxes are described applying the Monin-Obukhov similarity theory, where the flux depends on the air-surface temperature difference and wind speed. The theory makes idealized assumptions and the resulting estimates often have large errors. This is the case particularly in conditions when the air is warmer than the Earth surface, i.e., the atmospheric boundary layer is stably stratified, and turbulence is therefore weak. This is a common situation over snow and ice in the Arctic and Antarctic. In this paper, we present alternative models for heat flux estimation evolved by means of genetic programming (GP). To this aim, we utilize the best heat flux data collected in the Arctic and Antarctic sea ice zones. We obtain GP models that are more accurate, robust, and conceptually novel from the viewpoint of meteorology. Contrary to the Monin-Obukhov theory, the GP equations are not solely based on the air-surface temperature difference and wind speed, but include also radiative fluxes that improve the performance of the method. These results open the door to a new class of approaches to heat flux prediction with potential applications in weather and climate models.
BibTeX:
@inproceedings{Stanislawska2015,
  author = {Stanislawska, Karolina and Krawiec, Krzysztof and Vihma, Timo},
  title = {Genetic Programming for Estimation of Heat Flux Between the Atmosphere and Sea Ice in Polar Regions},
  booktitle = {Proceedings of the 2015 Annual Conference on Genetic and Evolutionary Computation},
  publisher = {ACM},
  year = {2015},
  pages = {1279--1286},
  url = {http://doi.acm.org/10.1145/2739480.2754675},
  doi = {http://doi.org/10.1145/2739480.2754675}
}
Swan, J., Adriaensen, S., Bishr, M., Burke, E.K., Clark, J.A., Causmaecker, P.D., Durillo, J., Hammond, K., Hart, E., Johnson, C.G., Kocsis, Z.A., Kovitz, B., Krawiec, K., Martin, S., Merelo, J., Minku, L.L., Ozcan, E., Pappa, G.L., Pesch, E., Garcıa-Sánchez, P., Schaerf, A., Sim, K., Smith, J.E., Stützle, T., Voß, S., Wagner, S. and Yao, X. A Research Agenda for Metaheuristic Standardization 2015 MIC 2015: The XI Metaheuristics International Conference  inproceedings  
Abstract: We propose that the development of standardized, explicit, machine-readable descriptions of metaheuris- tics will greatly advance scientific progress in the field. In particular, we advocate a purely functional description of metaheuristics — separate from any metaphors that inspire them and with no hidden mech- anisms. A recent policy statement in the Journal of Heuristics1 highlights the need for improved research practice via increased transparency of implementation and principled decomposition of metaheuristics to isolate the contribution of their component parts. We argue here that addressing these issues via explicit descriptions can also offer further benefits. Standardization and pure-functional descriptions promote a higher standard of rigor for both communication and reproducibility of results. The modularity of our proposed approach opens up opportunities to compose heuristics in novel ways, along with better support for parallel processing. Most significantly, it is the basis for a greater degree of mechanized reasoning: we discuss how this might support large-scale collaborative research activity, leading to automated dis- covery, mining and assembly of metaheuristics.
BibTeX:
@inproceedings{Swan2015:MIC,
  author = {Jerry Swan and Steven Adriaensen and Mohamed Bishr and Edmund K Burke and John A Clark and Patrick De Causmaecker and Juanjo Durillo and Kevin Hammond and Emma Hart and Colin G Johnson and Zoltan A Kocsis and Ben Kovitz and Krzysztof Krawiec and Simon Martin and JJ Merelo and Leandro L Minku and Ender Ozcan and Gisele L Pappa and Erwin Pesch and Pablo Garcıa-Sánchez and Andrea Schaerf and Kevin Sim and Jim E Smith and Thomas Stützle and Stefan Voß and Stefan Wagner and Xin Yao},
  title = {A Research Agenda for Metaheuristic Standardization},
  booktitle = {MIC 2015: The XI Metaheuristics International Conference},
  year = {2015}
}
Szubert, M., Jaśkowski, W., Liskowski, P. and Krawiec, K. The Role of Behavioral Diversity and Difficulty of Opponents in Coevolving Game-Playing Agents 2015
Vol. 9028Applications of Evolutionary Computation, pp. 394-405 
inproceedings DOI URL 
Abstract: Generalization performance of learning agents depends on the training experience to which they have been exposed. In game-playing domains, that experience is determined by the opponents faced during learning. This analytical study investigates two characteristics of opponents in competitive coevolutionary learning: behavioral diversity and difficulty (performance against other players). To assess diversity, we propose a generic intra-game behavioral distance measure, that could be adopted to other sequential decision problems. We monitor both characteristics in two-population coevolutionary learning of Othello strategies, attempting to explain their relationship with the generalization performance achieved by the evolved solutions. The main observation is the existence of a non-obvious trade-off between difficulty and diversity, with the latter being essential for obtaining high generalization performance.
BibTeX:
@inproceedings{SzubertEvoGames2015,
  author = {Szubert, Marcin and Jaśkowski, Wojciech and Liskowski, Paweł and Krawiec, Krzysztof},
  title = {The Role of Behavioral Diversity and Difficulty of Opponents in Coevolving Game-Playing Agents},
  booktitle = {Applications of Evolutionary Computation},
  publisher = {Springer International Publishing},
  year = {2015},
  volume = {9028},
  pages = {394-405},
  url = {http://dx.doi.org/10.1007/978-3-319-16549-3_32},
  doi = {http://doi.org/10.1007/978-3-319-16549-3_32}
}
Arnaldo, I., Krawiec, K. and O'Reilly, U.-M. Multiple Regression Genetic Programming 2014 Proceeding of the sixteenth annual conference on Genetic and evolutionary computation conference  inproceedings DOI URL 
Abstract: We propose a new means of executing a genetic program which improves
its output quality. Our approach, called Multiple Regression Genetic
Programming (MRGP) decouples and linearly combines a program's subexpressions
via multiple regression on the target variable. The regression yields
an alternate output: the prediction of the resulting multiple regression
model. It is this output, over many fitness cases, that we assess
for fitness, rather than the program's execution output. MRGP can
be used to improve the fitness of a final evolved solution. On our experimental suite,
MRGP consistently generated solutions fitter than the result of competent GP or multiple regression.
When integrated into GP, inline MRGP, on the basis of equivalent computational budget,
outperforms competent GP while also besting posthoc MRGP.
Thus MRGP's output method is shown to be superior to the output of
program execution and it represents a practical, cost neutral, improvement to GP.
BibTeX:
@inproceedings{ArnaldoKrawiecOReily:2014:GECCO,
  author = {Ignacio Arnaldo and Krzysztof Krawiec and Una-May O'Reilly},
  title = {Multiple Regression Genetic Programming},
  booktitle = {Proceeding of the sixteenth annual conference on Genetic and evolutionary computation conference},
  publisher = {ACM},
  year = {2014},
  url = {http://dx.doi.org/10.1145/2576768.2598291},
  doi = {dx.doi.org/10.1145/2576768.2598291}
}
Jaskowski, W., Krawiec, K. and Wieloch, B. Cross-Task Code Reuse in Genetic Programming Applied to Visual Learning 2014 International Journal of Applied Mathematics and Computer Science
Vol. 24(1), pp. 183–197  
article DOI URL 
Abstract: We propose a method that enables effective code reuse between evolutionary runs that solve a set of related visual learning tasks. We start with introducing a visual learning approach that uses genetic programming individuals to recognize objects. The process of recognition is generative, i.e., requires the learner to restore the shape of the processed object. This method is extended with a code reuse mechanism by introducing a crossbreeding operator that allows importing the genetic material from other evolutionary runs. In the experimental part, we compare the performance of the extended approach to the basic method on a real-world task of handwritten character recognition, and conclude that code reuse leads to better results in terms of fitness and recognition accuracy. The detailed analysis of the crossbred genetic material shows also that code reuse is most profitable when the recognized objects exhibit visual similarity.
BibTeX:
@article{2013AMCS,
  author = {Wojciech Jaskowski and Krzysztof Krawiec and Bartosz Wieloch},
  title = {Cross-Task Code Reuse in Genetic Programming Applied to Visual Learning},
  journal = {International Journal of Applied Mathematics and Computer Science},
  year = {2014},
  volume = {24},
  number = {1},
  pages = { 183–197 },
  url = {http://www.cs.put.poznan.pl/kkrawiec/pubs/2013AMCS.pdf},
  doi = {http://doi.org/10.2478/amcs-2014-0014}
}
Krawiec, K. Metaheuristic Design Pattern: Candidate Solution Repair 2014 Proceedings of the 2014 Conference Companion on Genetic and Evolutionary Computation Companion, pp. 1415-1418  inproceedings DOI URL 
Abstract:
In metaheuristic algorithms applied to certain problems, it may be difficult to design search operators that guarantee producing feasible search points. In such cases, it may be more efficient to allow a search operator to yield an infeasible solution, and then turn it into a feasible one using a repair process. This paper is an attempt to provide a broad perspective on the candidate solution repair and frame it as a metaheuristic design pattern.
BibTeX:
@inproceedings{Krawiec:2014:GECCOWorkshop,
  author = {Krawiec, Krzysztof},
  title = {Metaheuristic Design Pattern: Candidate Solution Repair},
  booktitle = {Proceedings of the 2014 Conference Companion on Genetic and Evolutionary Computation Companion},
  publisher = {ACM},
  year = {2014},
  pages = {1415--1418},
  url = {http://doi.acm.org/10.1145/2598394.2609847},
  doi = {http://doi.org/10.1145/2598394.2609847}
}
Krawiec, K. and O'Reilly, U.-M. Behavioral Search Drivers for Genetic Programing 2014
Vol. 8599Genetic Programming, pp. 210-221 
inproceedings  
Abstract: Synthesizing a program with the desired input-output behavior by means of genetic programming is an iterative process that needs appropriate guidance. That guidance is conventionally provided by a fitness function that measures the conformance of program output with the desired output. Contrary to widely adopted stance, there is no evidence that this quality measure is the best choice; alternative search drivers may exist that make search more effective. This study proposes and investigates a new family of behavioral search drivers, which inspect not only final program output, but also program behavior meant as the partial results it arrives at while executed.
BibTeX:
@inproceedings{krawiec2:2014:EuroGP,
  author = {Krzysztof Krawiec and Una-May O'Reilly},
  title = {Behavioral Search Drivers for Genetic Programing},
  booktitle = {Genetic Programming},
  publisher = {Springer},
  year = {2014},
  volume = {8599},
  pages = {210--221}
}
Krawiec, K. and O'Reilly, U.-M. Behavioral programming: a broader and more detailed take on semantic GP 2014 GECCO '14: Proceedings of the 2014 conference on Genetic and evolutionary computation, pp. 935-942  inproceedings DOI URL 
Abstract: In evolutionary computation, the fitness of a
candidate solution conveys sparse feedback. Yet in many
cases, candidate solutions can potentially yield more
information. In genetic programming (GP), one can
easily examine program behaviour on particular fitness
cases or at intermediate execution states. However, how
to exploit it to effectively guide the search remains
unclear. In this study we apply machine learning
algorithms to features describing the intermediate
behavior of the executed program. We then drive the
standard evolutionary search with additional objectives
reflecting this intermediate behavior. The machine
learning functions independent of task-specific
knowledge and discovers potentially useful components
of solutions (subprograms), which we preserve in an
archive and use as building blocks when composing new
candidate solutions. In an experimental assessment on a
suite of benchmarks, the proposed approach proves more
capable of finding optimal and/or well-performing
solutions than control methods.
BibTeX:
@inproceedings{Krawiec:2014:GECCO-short,
  author = {Krzysztof Krawiec and Una-May O'Reilly},
  title = {Behavioral programming: a broader and more detailed take on semantic GP},
  booktitle = {GECCO '14: Proceedings of the 2014 conference on Genetic and evolutionary computation},
  publisher = {ACM},
  year = {2014},
  pages = {935--942},
  url = {http://doi.acm.org/10.1145/2576768.2598288},
  doi = {http://doi.org/10.1145/2576768.2598288}
}
Krawiec, K. and O'Reilly, U.-M. Behavioral Programming: A Broader and More Detailed Take on Semantic GP 2014 Proceeding of the sixteenth annual conference on Genetic and evolutionary computation conference  inproceedings DOI URL 
Abstract: In evolutionary computation, the fitness of a candidate solutionfunction conveys sparse feedback. Yet in many cases, candidate solutions can potentially yield more information. In genetic programming (GP), one can easily examine program behavior on particular fitness cases or at intermediate execution states. However, how to exploit it to effectively guide the search remains unclear. In this study we apply machine learning algorithms to features describing the intermediate behavior of the executed program. We then drive the standard evolutionary search with additional objectives reflecting this intermediate behavior. The machine learning functions independent of task-specific knowledge and discovers potentially useful components of solutions (subprograms), which we preserve in an archive and use as building blocks when composing new candidate solutions. In an experimental assessment on a suite of benchmarks, the proposed approach proves more capable of finding optimal and/or well-performing solutions than control methods.
BibTeX:
@inproceedings{KrawiecOReily:2014:GECCO,
  author = {Krzysztof Krawiec and Una-May O'Reilly},
  title = {Behavioral Programming: A Broader and More Detailed Take on Semantic GP},
  booktitle = {Proceeding of the sixteenth annual conference on Genetic and evolutionary computation conference},
  publisher = {ACM},
  year = {2014},
  url = {http://dx.doi.org/10.1145/2576768.2598288},
  doi = {dx.doi.org/10.1145/2576768.2598288}
}
Krawiec, K. and Solar-Lezama, A. Improving Genetic Programming with Behavioral Consistency Measure 2014
Vol. 8672Parallel Problem Solving from Nature -- PPSN XIII, pp. 434-443 
inproceedings DOI  
Abstract: Program synthesis tasks usually specify only the desired output of a program
and do not state any expectations about its internal behavior. The intermediate execution
states reached by a running program can be nonetheless deemed as more or less preferred according
to their information content with respect to the desired output. In this paper,
a consistency measure is proposed that implements this observation. When used as an
additional search objective in a typical genetic programming setting,
this measure improves the success rate on a suite of 35 benchmarks
in a statistically significant way.
BibTeX:
@inproceedings{LNCS86720434,
  author = {Krzysztof Krawiec and Armando Solar-Lezama},
  title = {Improving Genetic Programming with Behavioral Consistency Measure},
  booktitle = {Parallel Problem Solving from Nature -- PPSN XIII},
  publisher = {Springer},
  year = {2014},
  volume = {8672},
  pages = {434--443},
  doi = {http://doi.org/10.1007/978-3-319-10762-2_43}
}
Liskowski, P. and Krawiec, K. Discovery of Implicit Objectives by Compression of Interaction Matrix in Test-Based Problems 2014
Vol. 8672Parallel Problem Solving from Nature -- PPSN XIII, pp. 611-620 
inproceedings DOI  
Abstract: In test-based problems, commonly solved with competitive coevolution
algorithms, candidate solutions (e.g., game strategies) are evaluated
by interacting with tests (e.g., opponents). As the number of tests
is typically large, it is expensive to calculate the exact value of
objective function, and one has to elicit a useful training signal
(search gradient) from the outcomes of a limited number of interactions
between these coevolving entities. Averaging of interaction outcomes,
typically used to that aim, ignores the fact that solutions often
have to master different and unrelated skills, which form underlying
objectives of the problem. We propose a method for on-line discovery
of such objectives via heuristic compression of interaction outcomes.
The compressed matrix implicitly defines derived search objectives
that can be used by traditional multiobjective search techniques
(NSGA-II in this study). When applied to the challenging variant of
multi-choice Iterated Prisoner's Dilemma problem, the proposed approach
outperforms conventional two-population coevolution in a statistically
significant way.
BibTeX:
@inproceedings{LNCS86720611,
  author = {Paweł Liskowski and Krzysztof Krawiec},
  title = {Discovery of Implicit Objectives by Compression of Interaction Matrix in Test-Based Problems},
  booktitle = {Parallel Problem Solving from Nature -- PPSN XIII},
  publisher = {Springer},
  year = {2014},
  volume = {8672},
  pages = {611--620},
  doi = {http://doi.org/10.1007/978-3-319-10762-2_60}
}
Ryan, C., Krawiec, K., O'Reilly, U.-M., Fitzgerald, J. and Medernach, D. Building a Stage 1 Computer Aided Detector for Breast Cancer using Genetic Programming 2014
Vol. 8599Genetic Programming, pp. 162-173 
inproceedings DOI URL 
Abstract: We describe a fully automated workflow for performing stage 1 breast cancer detection with GP as its cornerstone. Mammograms are by far the most widely used method for detecting breast cancer in women, and its use in national screening can have a dramatic impact on early detection and survival rates. With the increased availability of digital mammography, it is becoming increasingly more feasible to use automated methods to help with detection.
A stage 1 detector examines mammograms and highlights suspicious areas that require further investigation. A too conservative approach degenerates to marking every mammogram (or segment of) as suspicious, while missing a cancerous area can be disastrous.
Our workflow positions us right at the data collection phase such that we generate textural features ourselves. These are fed through our system, which performs PCA on them before passing the most salient ones to GP to generate classifiers. The classifiers give results of 100% accuracy on true positives and a false positive per image rating of just 1.5, which is better than prior work. Not only this, but our system can use GP as part of a feedback loop, to both select and help generate further features.
BibTeX:
@inproceedings{krawiec:2014:EuroGP,
  author = {Conor Ryan and Krzysztof Krawiec and Una-May O'Reilly and Jeannie Fitzgerald and David Medernach},
  title = {Building a Stage 1 Computer Aided Detector for Breast Cancer using Genetic Programming},
  booktitle = {Genetic Programming},
  publisher = {Springer Berlin Heidelberg},
  year = {2014},
  volume = {8599},
  pages = {162-173},
  url = {http://dx.doi.org/10.1007/978-3-662-44303-3_14},
  doi = {http://doi.org/10.1007/978-3-662-44303-3_14}
}
Ryan, C., Krawiec, K., O'Reilly, U.-M., Fitzgerald, J. and Medernach, D. Building a Stage 1 Computer Aided Detector for Breast Cancer using Genetic Programming 2014
Vol. 859917th European Conference on Genetic Programming, pp. 162-173 
inproceedings DOI  
Abstract: We describe a fully automated workflow for performing
stage1 breast cancer detection with GP as its
cornerstone. Mammograms are by far the most widely used
method for detecting breast cancer in women, and its
use in national screening can have a dramatic impact on
early detection and survival rates. With the increased
availability of digital mammography, it is becoming
increasingly more feasible to use automated methods to
help with detection. A stage 1 detector examines
mammograms and highlights suspicious areas that require
further investigation. A too conservative approach
degenerates to marking every mammogram (or segment of)
as suspicious, while missing a cancerous area can be
disastrous. Our workflow positions us right at the data
collection phase such that we generate textural
features ourselves. These are fed through our system,
which performs PCA on them before passing the most
salient ones to GP to generate classifiers. The
classifiers give results of 100percent accuracy on true
positives and a false positive per image rating of just
1.5, which is better than prior work. Not only this,
but our system can use GP as part of a feedback loop,
to both select and help generate further features.
BibTeX:
@inproceedings{ryan:2014:EuroGP-short,
  author = {Conor Ryan and Krzysztof Krawiec and Una-May O'Reilly and Jeannie Fitzgerald and David Medernach},
  title = {Building a Stage 1 Computer Aided Detector for Breast Cancer using Genetic Programming},
  booktitle = {17th European Conference on Genetic Programming},
  publisher = {Springer},
  year = {2014},
  volume = {8599},
  pages = {162--173},
  doi = {http://doi.org/10.1007/978-3-662-44303-3_14}
}
Swan, J., Drake, J. and Krawiec, K. Substituting Closed-Form Expressions for ERCs 2014   unpublished  
BibTeX:
@unpublished{SwanDrakeKrawiec:2014:SMGP,
  author = {Jerry Swan and John Drake and Krzysztof Krawiec},
  title = {Substituting Closed-Form Expressions for ERCs},
  year = {2014},
  note = {Presented at Semantic Methods in Genetic Programming, PPSN'14 Workshop}
}
Swan, J., Neumann, G.K. and Krawiec, K. Analysis of Semantic Building Blocks via Gröbner Bases 2014   unpublished  
BibTeX:
@unpublished{SwanNeumannKrawiec:2014:SMGP,
  author = {Jerry Swan and Geoffrey K. Neumann and Krzysztof Krawiec},
  title = {Analysis of Semantic Building Blocks via Gröbner Bases},
  year = {2014},
  note = {Presented at Semantic Methods in Genetic Programming, PPSN'14 Workshop}
}
Tomasz P. Pawlak, K.K. Guarantees of Progress for Geometric Semantic Genetic Programming 2014   unpublished  
BibTeX:
@unpublished{PawlakKrawiec:2014:SMGP,
  author = {Tomasz P. Pawlak, Krzysztof Krawiec},
  title = {Guarantees of Progress for Geometric Semantic Genetic Programming},
  year = {2014},
  note = {Presented at Semantic Methods in Genetic Programming, PPSN'14 Workshop}
}
Proceedings of the 17th European Conference on Genetic Programming, EuroGP 2014 2014
Vol. 8599 
proceedings DOI URL 
BibTeX:
@proceedings{LNCS8599,,
  title = {Proceedings of the 17th European Conference on Genetic Programming, EuroGP 2014},
  publisher = {Springer Verlag},
  year = {2014},
  volume = {8599},
  url = {http://www.evostar.org/EvoStar14confHandbook.pdf},
  doi = {http://doi.org/10.1007/978-3-662-44303-3}
}
Jaśkowski, W., Liskowski, P., Szubert, M. and Krawiec, K. Improving coevolution by random sampling 2013 Proceeding of the fifteenth annual conference on Genetic and evolutionary computation conference, pp. 1141-1148  inproceedings DOI URL 
Abstract: Recent developments cast doubts on the effectiveness of coevolutionary learning in interactive domains. A simple evolution with fitness evaluation based on games with random strategies has been found to generalize better than competitive coevolution. In an attempt to investigate this phenomenon, we analyze the utility of random opponents for one and two-population competitive coevolution applied to learning strategies for the game of Othello. We show that if coevolution uses two-population setup and engages also random opponents, it is capable of producing equally good strategies as evolution with random sampling for the expected utility performance measure. To investigate the differences between analyzed methods, we introduce performance profile, a tool that measures the player's performance against opponents of various strength. The profiles reveal that evolution with random sampling produces players coping well with mediocre opponents, but playing relatively poorly against stronger ones. This finding explains why in the round-robin tournament, evolution with random sampling is one of the worst methods from all those considered in this study.
BibTeX:
@inproceedings{Jaskowski:2013:ICR,
  author = {Jaśkowski, Wojciech and Liskowski, Paweł and Szubert, Marcin and Krawiec, Krzysztof},
  title = {Improving coevolution by random sampling},
  booktitle = {Proceeding of the fifteenth annual conference on Genetic and evolutionary computation conference},
  publisher = {ACM},
  year = {2013},
  pages = {1141--1148},
  url = {http://doi.acm.org/10.1145/2463372.2463512},
  doi = {http://doi.org/10.1145/2463372.2463512}
}
Krawiec, K. Genetic programming: where meaning emerges from program code 2013 Genetic Programming and Evolvable Machines, pp. 1-3  article DOI URL 
Abstract: Program behavior results from the interactions of instructions with data. In genetic programming, a substantial part of that behavior is not explicitly rewarded by fitness function, and thus emergent. This includes the intermediate memory states traversed by the executing programs. We argue that the potentially useful intermediate states can be detected and used to make evolutionary search more effective.
BibTeX:
@article{krawiec:2013:GPEM,
  author = {Krawiec, Krzysztof},
  title = {Genetic programming: where meaning emerges from program code},
  journal = {Genetic Programming and Evolvable Machines},
  publisher = {Springer US},
  year = {2013},
  pages = {1-3},
  url = {http://dx.doi.org/10.1007/s10710-013-9200-2},
  doi = {http://doi.org/10.1007/s10710-013-9200-2}
}
Krawiec, K. and Nawrocki, M. Implicit Fitness Sharing for Evolutionary Synthesis of License Plate Detectors 2013
Vol. 7835Applications of Evolutionary Computing, EvoApplications 2012, pp. 376-386 
inproceedings DOI  
Abstract: A genetic programming algorithm for synthesis of
object detection systems is proposed and applied to the
task of license plate recognition in uncontrolled
lighting conditions. The method evolves solutions
represented as data flows of high-level parametric
image operators. In an extended variant, the algorithm
employs implicit fitness sharing, which allows
identifying the particularly difficult training
examples and focusing the training process on them. The
experiment, involving heterogeneous video sequences
acquired in diverse conditions, demonstrates that
implicit fitness sharing substantially improves the
predictive performance of evolved detection systems,
providing maximum recognition accuracy achievable for
the considered setup and training data.
BibTeX:
@inproceedings{Krawiec:2013:EvoIASP-short,
  author = {Krzysztof Krawiec and Mateusz Nawrocki},
  title = {Implicit Fitness Sharing for Evolutionary Synthesis of License Plate Detectors},
  booktitle = {Applications of Evolutionary Computing, EvoApplications 2012},
  publisher = {Springer},
  year = {2013},
  volume = {7835},
  pages = {376--386},
  doi = {http://doi.org/10.1007/978-3-642-37192-9_38}
}
Krawiec, K. and Nawrocki, M. Implicit Fitness Sharing for Evolutionary Synthesis of License Plate Detectors 2013
Vol. 7835Applications of Evolutionary Computation, pp. 376-386 
inproceedings  
BibTeX:
@inproceedings{LNCS78350376,
  author = {Krzysztof Krawiec and Mateusz Nawrocki},
  title = {Implicit Fitness Sharing for Evolutionary Synthesis of License Plate Detectors},
  booktitle = {Applications of Evolutionary Computation},
  publisher = {Springer},
  year = {2013},
  volume = {7835},
  pages = {376--386}
}
Krawiec, K. and Pawlak, T. Locally Geometric Semantic Crossover: A Study on the Roles of Semantics and Homology in Recombination Operators 2013 Genetic Programming and Evolvable Machines
Vol. 14(1), pp. 31-63 
article DOI URL 
BibTeX:
@article{Krawiec:2013:LGS,
  author = {Krawiec, Krzysztof and Pawlak, Tomasz},
  title = {Locally Geometric Semantic Crossover: A Study on the Roles of Semantics and Homology in Recombination Operators},
  journal = {Genetic Programming and Evolvable Machines},
  publisher = {Kluwer Academic Publishers},
  year = {2013},
  volume = {14},
  number = {1},
  pages = {31--63},
  url = {http://dx.doi.org/10.1007/s10710-012-9172-7},
  doi = {http://doi.org/10.1007/s10710-012-9172-7}
}
Krawiec, K. and Pawlak, T. Locally geometric semantic crossover: a study on the roles of semantics and homology in recombination operators 2013 Genetic Programming and Evolvable Machines
Vol. 14(1), pp. 31-63 
article DOI URL 
Abstract: This study presents an extensive account of Locally Geometric Semantic Crossover (LGX), a semantically-aware recombination operator for genetic programming (GP). LGX is designed to exploit the semantic properties of programs and subprograms, in particular the geometry of semantic space that results from distance-based fitness functions used predominantly in GP. When applied to a pair of parents, LGX picks in them at random a structurally common (homologous) locus, calculates the semantics of subprograms located at that locus, finds a procedure that is semantically medial with respect to these subprograms, and replaces them with that procedure. The library of procedures is prepared prior to the evolutionary run and indexed by a multidimensional structure (kd-tree) allowing for efficient search. The paper presents the rationale for LGX design and an extensive computational experiment concerning performance, computational cost, impact on program size, and capability of generalization. LGX is compared with six other operators, including conventional tree-swapping crossover, semantic-aware operators proposed in previous studies, and control methods designed to verify the importance of homology and geometry of the semantic space. The overall conclusion is that LGX, thanks to combination of the semantically medial operation with homology, improves the efficiency of evolutionary search, lowers the variance of performance, and tends to be more resistant to overfitting.
BibTeX:
@article{springerlink:10.1007/s10710-012-9172-7,
  author = {Krawiec, Krzysztof and Pawlak, Tomasz},
  title = {Locally geometric semantic crossover: a study on the roles of semantics and homology in recombination operators},
  journal = {Genetic Programming and Evolvable Machines},
  publisher = {Springer US},
  year = {2013},
  volume = {14},
  number = {1},
  pages = {31-63},
  url = {http://dx.doi.org/10.1007/s10710-012-9172-7},
  doi = {http://doi.org/10.1007/s10710-012-9172-7}
}
Krawiec, K. and Pawlak, T. Approximating geometric crossover by semantic backpropagation 2013 Proceeding of the fifteenth annual conference on Genetic and evolutionary computation conference, pp. 941-948  inproceedings DOI URL 
Abstract: We propose a novel crossover operator for tree-based genetic programming, that produces approximately geometric offspring. We empirically analyze certain aspects of geometry of crossover operators and verify performance of the new operator on both, training and test fitness cases coming from set of symbolic regression benchmarks. The operator shows superior performance and higher probability of producing geometric offspring than tree-swapping crossover and other semantic-aware control methods.
BibTeX:
@inproceedings{Krawiec:2013:AGC:2463372.2463483,
  author = {Krawiec, Krzysztof and Pawlak, Tomasz},
  title = {Approximating geometric crossover by semantic backpropagation},
  booktitle = {Proceeding of the fifteenth annual conference on Genetic and evolutionary computation conference},
  publisher = {ACM},
  year = {2013},
  pages = {941--948},
  url = {http://doi.acm.org/10.1145/2463372.2463483},
  doi = {http://doi.org/10.1145/2463372.2463483}
}
Krawiec, K. and Swan, J. Pattern-guided genetic programming 2013 Proceeding of the fifteenth annual conference on Genetic and evolutionary computation conference, pp. 949-956  inproceedings DOI URL 
Abstract: Online progress in search and optimization is often hindered by neutrality in the fitness landscape, when many genotypes map to the same fitness value. We propose a method for imposing a gradient on the fitness function of a metaheuristic (in this case, Genetic Programming) via a metric (Minimum Description Length) induced from patterns detected in the trajectory of program execution. These patterns are induced via a decision tree classifier. We apply this method to a range of integer and boolean-valued problems, significantly outperforming the standard approach. The method is conceptually straightforward and applicable to virtually any metaheuristic that can be appropriately instrumented.
BibTeX:
@inproceedings{Krawiec:2013:PGP:2463372.2463496,
  author = {Krawiec, Krzysztof and Swan, Jerry},
  title = {Pattern-guided genetic programming},
  booktitle = {Proceeding of the fifteenth annual conference on Genetic and evolutionary computation conference},
  publisher = {ACM},
  year = {2013},
  pages = {949--956},
  url = {http://doi.acm.org/10.1145/2463372.2463496},
  doi = {http://doi.org/10.1145/2463372.2463496}
}
Krawiec, K. and Swan, J. Guiding Evolutionary Learning by Searching for Regularities in Behavioral Trajectories: A Case for Representation Agnosticism 2013 How Should Intelligence be Abstracted in AI Research: MDPs, Symbolic Representations, Artificial Neural Networks, or ...? AAAI Fall Symposium 2013, pp. 41-46  inproceedings URL 
Abstract: An intelligent agent can display behavior that is not directly related to the task it learns. Depending on the adopted AI framework and task formulation, such behavior is sometimes attributed to environment exploration, or ignored as irrelevant, or even penalized as undesired. We postulate here that virtually every interaction of an agent with its learning environment can result in outcomes that carry information which can be potentially exploited to solve the task. To support this claim, we present Pattern Guided Evolutionary Algorithm (PANGEA), an extension of genetic programming (GP), a genre of evolutionary computation that aims at synthesizing programs that display the desired input-output behavior. PANGEA uses machine learning to search for regularities in intermediate outcomes of program execution (which are ignored in standard GP), more specifically for relationships between these outcomes and the desired program output. The information elicited in this way is used to guide the evolutionary learning process by appropriately adjusting program fitness. An experiment conducted on a suite of benchmarks demonstrates that this architecture makes agent learning more effective than in conventional GP. In the paper, we discuss the possible generalizations and extensions of this architecture and its relationships with other contemporary paradigms like novelty search and deep learning. In conclusion, we extrapolate PANGEA to postulate a dynamic and behavioral learning framework for intelligent agents.
BibTeX:
@inproceedings{krawiec:2013:AAAI,
  author = {Krawiec, Krzysztof and Swan, Jerry},
  title = {Guiding Evolutionary Learning by Searching for Regularities in Behavioral Trajectories: A Case for Representation Agnosticism},
  booktitle = {How Should Intelligence be Abstracted in AI Research: MDPs, Symbolic Representations, Artificial Neural Networks, or ...? AAAI Fall Symposium 2013},
  publisher = {AAAI},
  year = {2013},
  pages = {41--46},
  url = {http://www.cs.put.poznan.pl/kkrawiec/pubs/FS13-02-013.pdf}
}
Sikorski, B.L., Bukowska, D., Rumiński, D., Gorczyńska, I., Szkulmowski, M., Krawiec, K., Malukiewicz, G. and Wojtkowski, M. Visualization of 3D retinal microcapillary network using OCT 2013 Acta Ophthalmologica
Vol. 91, pp. 0-0 
article DOI URL 
BibTeX:
@article{AOS:AOS4466,
  author = {Sikorski, Bartosz L. and Bukowska, Danuta and Rumiński, Daniel and Gorczyńska, Iwona and Szkulmowski, Maciej and Krawiec, Krzysztof and Malukiewicz, Grażyna and Wojtkowski, Maciej},
  title = {Visualization of 3D retinal microcapillary network using OCT},
  journal = {Acta Ophthalmologica},
  publisher = {Blackwell Publishing Ltd},
  year = {2013},
  volume = {91},
  pages = {0--0},
  url = {http://onlinelibrary.wiley.com/doi/10.1111/j.1755-3768.2013.4466.x/abstract},
  doi = {http://doi.org/10.1111/j.1755-3768.2013.4466.x}
}
Stanislawska, K., Kundzewicz, Z.W. and Krawiec, K. Hindcasting global temperature by evolutionary computation 2013 Acta Geophysica
Vol. 61(3) 
article  
Abstract: Interpretation of changes of global temperature is important for our understanding of the climate system and for our confidence in projections for the future. Massive efforts have been devoted to improve the accuracy of reproducing the global temperature by the available climate models, but the hindcasts are still inaccurate. Notwithstanding the need to further advance climate models, one may consider data-driven approaches, providing practically useful results in a simpler and faster way. Without assuming any prior knowledge about physics and without imposing a model structure that encapsulates the existing knowledge about the underlying processes, we hindcast global temperature by automatically identified evolutionary computation (genetic programming) models. We use 60 years of records of global temperature and climate drivers, with training and testing periods being 1950-1999 and 2000-2009, respectively. This paper demonstrates that the global temperature observed in the past is mimicked with reasonably good accuracy. Evolutionary computation holds promise for modeling the global climate system, which looks hopelessly complex in classical perspective.
BibTeX:
@article{Stanislawska2013,
  author = {Karolina Stanislawska and Zbigniew W. Kundzewicz and Krzysztof Krawiec},
  title = {Hindcasting global temperature by evolutionary computation},
  journal = {Acta Geophysica},
  year = {2013},
  volume = {61},
  number = {3}
}
Szubert, M., Jaśkowski, W. and Krawiec, K. On Scalability, Generalization, and Hybridization of Coevolutionary Learning: A Case Study for Othello 2013 Computational Intelligence and AI in Games, IEEE Transactions on
Vol. 5(3), pp. 214-226 
article DOI  
Abstract: This study investigates different methods of learning to play the game of Othello. The main questions posed concern scalability of algorithms with respect to the search space size and their capability to generalize and produce players that fare well against various opponents. The considered algorithms represent strategies as n-tuple networks, and employ self-play temporal difference learning (TDL), evolutionary learning (EL) and coevolutionary learning (CEL), and hybrids thereof. To assess the performance, three different measures are used: score against an a priori given opponent (a fixed heuristic strategy), against opponents trained by other methods (round-robin tournament), and against the top-ranked players from the online Othello League. We demonstrate that although evolutionary-based methods yield players that fare best against a fixed heuristic player, it is the coevolutionary temporal difference learning (CTDL), a hybrid of coevolution and TDL, that generalizes better and proves superior when confronted with a pool of previously unseen opponents. Moreover, CTDL scales well with the size of representation, attaining better results for larger n-tuple networks. By showing that a strategy learned in this way wins against the top entries from the Othello League, we conclude that it is one of the best 1-ply Othello players obtained to date without explicit use of human knowledge.
BibTeX:
@article{2013TCIAIGKrawiec,
  author = {Szubert, Marcin and Jaśkowski, Wojciech and Krawiec, Krzysztof},
  title = {On Scalability, Generalization, and Hybridization of Coevolutionary Learning: A Case Study for Othello},
  journal = {Computational Intelligence and AI in Games, IEEE Transactions on},
  year = {2013},
  volume = {5},
  number = {3},
  pages = {214-226},
  doi = {http://doi.org/10.1109/TCIAIG.2013.2258919}
}
Szubert, M., Jaśkowski, W., Liskowski, P. and Krawiec, K. Shaping fitness function for evolutionary learning of game strategies 2013 Proceeding of the fifteenth annual conference on Genetic and evolutionary computation conference, pp. 1149-1156  inproceedings DOI URL 
Abstract: In evolutionary learning of game-playing strategies, fitness evaluation is based on playing games with certain opponents. In this paper we investigate how the performance of these opponents and the way they are chosen influence the efficiency of learning. For this purpose we introduce a simple method for shaping the fitness function by sampling the opponents from a biased performance distribution. We compare the shaped function with existing fitness evaluation approaches that sample the opponents from an unbiased performance distribution or from a coevolving population. In an extensive computational experiment we employ these methods to learn Othello strategies and assess both the absolute and relative performance of the elaborated players. The results demonstrate the superiority of the shaping approach, and can be explained by means of performance profiles, an analytical tool that evaluate the evolved strategies using a range of variably skilled opponents.
BibTeX:
@inproceedings{Szubert:2013:SFF:2463372.2463513,
  author = {Szubert, Marcin and Jaśkowski, Wojciech and Liskowski, Paweł and Krawiec, Krzysztof},
  title = {Shaping fitness function for evolutionary learning of game strategies},
  booktitle = {Proceeding of the fifteenth annual conference on Genetic and evolutionary computation conference},
  publisher = {ACM},
  year = {2013},
  pages = {1149--1156},
  url = {http://doi.acm.org/10.1145/2463372.2463513},
  doi = {http://doi.org/10.1145/2463372.2463513}
}
Wieloch, B. and Krawiec, K. Running programs backwards: instruction inversion for effective search in semantic spaces 2013 Proceeding of the fifteenth annual conference on Genetic and evolutionary computation conference, pp. 1013-1020  inproceedings DOI URL 
Abstract: The instructions used for solving typical genetic programming tasks have strong mathematical properties. In this study, we leverage one of such properties: invertibility. A search operator is proposed that performs an approximate reverse execution of program fragments, trying to determine in this way the desired semantics (partial outcome) at intermediate stages of program execution. The desired semantics determined in this way guides the choice of a subprogram that replaces the old program fragment. An extensive computational experiment on 20 symbolic regression and Boolean domain problems leads to statistically significant evidence that the proposed Random Desired Operator outperforms all typical combinations of conventional mutation and crossover operators.
BibTeX:
@inproceedings{Wieloch:2013:RPB:2463372.2463493,
  author = {Wieloch, Bartosz and Krawiec, Krzysztof},
  title = {Running programs backwards: instruction inversion for effective search in semantic spaces},
  booktitle = {Proceeding of the fifteenth annual conference on Genetic and evolutionary computation conference},
  publisher = {ACM},
  year = {2013},
  pages = {1013--1020},
  url = {http://doi.acm.org/10.1145/2463372.2463493},
  doi = {http://doi.org/10.1145/2463372.2463493}
}
Genetic Programming 2013
Vol. 7831 
proceedings  
BibTeX:
@proceedings{LNCS7831,,
  title = {Genetic Programming},
  publisher = {Springer},
  year = {2013},
  volume = {7831}
}
Krawiec, K. On Relationships Between Semantic Diversity, Complexity and Modularity of Programming Tasks 2012 Proceedings of the Fourteenth International Conference on Genetic and Evolutionary Computation Conference, pp. 783-790  inproceedings DOI URL 
Abstract: We investigate semantic properties of linear programs, both internally,
by analyzing the memory states they produce during execution, and
externally, by inspecting program outcomes. The main concept of the
formalism we propose is program trace, which reflects the behavior
of program in semantic space. It allows us to characterize programming
tasks in terms of traces of programs that solve them, and to propose
certain measures that reveal their properties. We are primarily interested
in measures that quantitatively characterize functional (semantic,
behavioral) modularity of programming tasks. The experiments conducted
on large samples of linear programs written in Push demonstrate that
semantic structure varies from task to task, and reveal patterns
of different forms of modularity. In particular, we identify interesting
relationships between task modularity, task complexity, and program
length, and conclude that a great share of programming tasks are
modular.
BibTeX:
@inproceedings{krawiec:2012:GECCOPush,
  author = {Krawiec, Krzysztof},
  title = {On Relationships Between Semantic Diversity, Complexity and Modularity of Programming Tasks},
  booktitle = {Proceedings of the Fourteenth International Conference on Genetic and Evolutionary Computation Conference},
  publisher = {ACM},
  year = {2012},
  pages = {783--790},
  url = {http://doi.acm.org/10.1145/2330163.2330272},
  doi = {http://doi.org/10.1145/2330163.2330272}
}
Krawiec, K. Medial Crossovers for Genetic Programming 2012
Vol. 7244Genetic Programming, pp. 61-72 
inproceedings  
Abstract: We propose a class of crossover operators for genetic programming
that aim at making offspring programs semantically intermediate (medial)
with respect to parent programs by modifying short fragments of code
(subprograms). The approach is applicable to problems that define
fitness as a distance between program output and the desired output.
Based on that metric, we define two measures of semantic `mediality',
which we employ to design two crossover operators: one aimed at making
the semantic of offsprings geometric with respect to the semantic
of parents, and the other aimed at making them equidistant to parents'
semantics. The operators act only on randomly selected fragments
of parents' code, which makes them computationally efficient. When
compared experimentally with four other crossover operators, both
operators lead to success ratio at least as good as for the non-semantic
crossovers, and the operator based on equidistance proves superior
to all others.
BibTeX:
@inproceedings{krawiec:2012:EuroGP,
  author = {Krzysztof Krawiec},
  title = {Medial Crossovers for Genetic Programming},
  booktitle = {Genetic Programming},
  publisher = {Springer},
  year = {2012},
  volume = {7244},
  pages = {61--72}
}
Krawiec, K. and Pawlak, T. Quantitative Analysis of Locally Geometric Semantic Crossover 2012
Vol. 7491Parallel Problem Solving from Nature - PPSN XII, pp. 397-406 
inproceedings  
BibTeX:
@inproceedings{LNCS74910397,
  author = {Krzysztof Krawiec and Tomasz Pawlak},
  title = {Quantitative Analysis of Locally Geometric Semantic Crossover},
  booktitle = {Parallel Problem Solving from Nature - PPSN XII},
  publisher = {Springer},
  year = {2012},
  volume = {7491},
  pages = {397--406}
}
McDermott, J., White, D.R., Luke, S., Manzoni, L., Castelli, M., Vanneschi, L., Jaśkowski, W., Krawiec, K., Harper, R., De Jong, K. and O'Reilly, U.-M. Genetic programming needs better benchmarks 2012 Proceedings of the fourteenth international conference on Genetic and evolutionary computation conference, pp. 791-798  inproceedings DOI URL 
Abstract: Genetic programming (GP) is not a field noted for the rigor of its
benchmarking. Some of its benchmark problems are popular purely through
historical contingency, and they can be criticized as too easy or
as providing misleading information concerning real-world performance,
but they persist largely because of inertia and the lack of good
alternatives. Even where the problems themselves are impeccable,
comparisons between studies are made more difficult by the lack of
standardization. We argue that the definition of standard benchmarks
is an essential step in the maturation of the field. We make several
contributions towards this goal. We motivate the development of a
benchmark suite and define its goals; we survey existing practice;
we enumerate many candidate benchmarks; we report progress on reference
implementations; and we set out a concrete plan for gathering feedback
from the GP community that would, if adopted, lead to a standard
set of benchmarks.
BibTeX:
@inproceedings{McDermott:2012:GPN:2330163.2330273,
  author = {McDermott, James and White, David R. and Luke, Sean and Manzoni, Luca and Castelli, Mauro and Vanneschi, Leonardo and Jaśkowski, Wojciech and Krawiec, Krzysztof and Harper, Robin and De Jong, Kenneth and O'Reilly, Una-May},
  title = {Genetic programming needs better benchmarks},
  booktitle = {Proceedings of the fourteenth international conference on Genetic and evolutionary computation conference},
  publisher = {ACM},
  year = {2012},
  pages = {791--798},
  url = {http://doi.acm.org/10.1145/2330163.2330273},
  doi = {http://doi.org/10.1145/2330163.2330273}
}
Moraglio, A., Krawiec, K. and Johnson, C.G. Geometric Semantic Genetic Programming 2012
Vol. 7491Parallel Problem Solving from Nature - PPSN XII, pp. 21-31 
inproceedings  
BibTeX:
@inproceedings{LNCS74910021,
  author = {Alberto Moraglio and Krzysztof Krawiec and Colin G. Johnson},
  title = {Geometric Semantic Genetic Programming},
  booktitle = {Parallel Problem Solving from Nature - PPSN XII},
  publisher = {Springer},
  year = {2012},
  volume = {7491},
  pages = {21--31}
}
Stanislawska, K., Krawiec, K. and Kundzewicz, Z.W. Modelling global temperature changes with genetic programming 2012 Computer & Mathematics with Applications
Vol. 64(12), pp. 3717-3728 
article DOI URL 
Abstract: We use genetic programming (GP), a variant of
evolutionary computation, to build interpretable models
of global mean temperature as a function of natural and
anthropogenic forcings. In contrast to the conventional
approach, which engages models that are
physically-based but very data-demanding and
computation-intense, the proposed method is a
data-driven randomised search algorithm capable of
inducing a model from moderate amount of training data
at reasonable computational cost. GP maintains a
population of models and recombines them iteratively to
improve their performance meant as an ability to
explain the training data. Each model is a multiple
input-single output arithmetic expression built of a
predefined set of elementary components. Inputs include
external climate forcings, such as solar activity,
volcanic eruptions, composition of the atmosphere
(greenhouse gas concentration and aerosols), and
indices of internal variability (oscillations in the
Ocean-Atmosphere system), while the output is the
large-scale temperature. We used the data from the
period 1900-1999 for training and the period 2000-2009
for testing, and employed two quality measures: mean
absolute error and correlation coefficient. The
experiment showed that the models evolved by GP are
capable to predict, based exclusively on
non-temperature data, the global temperature more
accurately than a reference approach known in the
literature.
BibTeX:
@article{Stanislawska2012,
  author = {Karolina Stanislawska and Krzysztof Krawiec and Zbigniew W. Kundzewicz},
  title = {Modelling global temperature changes with genetic programming},
  journal = {Computer & Mathematics with Applications},
  year = {2012},
  volume = {64},
  number = {12},
  pages = {3717--3728},
  url = {http://www.sciencedirect.com/science/article/pii/S0898122112001745},
  doi = {http://doi.org/10.1016/j.camwa.2012.02.049}
}
Szubert, M. and Krawiec, K. Autonomous Shaping via Coevolutionary Selection of Training Experience 2012
Vol. 7492Parallel Problem Solving from Nature - PPSN XII, pp. 215-224 
inproceedings  
BibTeX:
@inproceedings{LNCS74920215,
  author = {Marcin Szubert and Krzysztof Krawiec},
  title = {Autonomous Shaping via Coevolutionary Selection of Training Experience},
  booktitle = {Parallel Problem Solving from Nature - PPSN XII},
  publisher = {Springer},
  year = {2012},
  volume = {7492},
  pages = {215--224}
}
Genetic Programming 2012
Vol. 7244Genetic Programming 
proceedings URL 
BibTeX:
@proceedings{moraglioBook:2012:EuroGP,,
  title = {Genetic Programming},
  booktitle = {Genetic Programming},
  publisher = {Springer},
  year = {2012},
  volume = {7244},
  url = {http://www.springer.com/alert/urltracking.do?id=Lc3037aM9f5965Sa9b9bc8}
}
Jaśkowski, W. and Krawiec, K. Formal Analysis, Hardness, and Algorithms for Extracting Internal Structure of Test-Based Problems 2011 Evolutionary Computation
Vol. 19(4)Evolutionary Computation, pp. 639-671 
article DOI URL 
Abstract: Problems in which some elementary entities interact with each other
are common in computational intelligence. This scenario, typical
for co-evolving artificial-life agents, learning strategies for games,
and machine learning from examples, can be formalized as test-based
problem and conveniently embedded in the common conceptual framework
of coevolution. In test-based problems candidate solutions are evaluated
on a number of test cases (agents, opponents, examples). It has been
recently shown that every test of such problem can be regarded as
a separate objective, and the whole problem as a multi-objective
optimization. Research on reducing the number of objectives while
preserving the relations between candidate solutions and tests led
to the notions of underlying objectives and internal problem structure,
which can be formalized as coordinate system that spatially arranges
candidate solutions and tests. The coordinate system that spans the
minimal number of axes determines the so-called dimension of a problem
and, being an inherent property of every problem, is of particular
interest. In this study, we investigate in-depth the formalism of
coordinate system and its properties, relate them to properties of
partially ordered sets, and design an exact algorithm for finding
a minimal coordinate system. We also prove that this problem is NP-hard
and come up with a heuristic which is superior to the best algorithm
proposed so far. Finally, we apply the algorithms to three abstract
problems and demonstrate that the dimension of the problem is typically
much lower than the number of tests, and for some problems converges
to the intrinsic parameter of the problem -- its a priori dimension.
BibTeX:
@article{2012jaskowskiECJ,
  author = {Jaśkowski, Wojciech and Krawiec, Krzysztof},
  title = {Formal Analysis, Hardness, and Algorithms for Extracting Internal Structure of Test-Based Problems},
  booktitle = {Evolutionary Computation},
  journal = {Evolutionary Computation},
  publisher = {MIT Press},
  year = {2011},
  volume = {19},
  number = {4},
  pages = {639--671},
  url = {http://dx.doi.org/10.1162/EVCO_a_00046},
  doi = {http://doi.org/10.1162/EVCO%7B%5C_%7Da%7B%5C_%7D00046}
}
Krawiec, K. Semantically embedded genetic programming: automated design of abstract program representations 2011 Proceedings of the 13th Annual Conference on Genetic and Evolutionary Computation, pp. 1379-1386  inproceedings DOI URL 
Abstract: We propose an alternative program representation that relies on automatic
design of semantic-based embeddings of genetic programs into discrete
multidimensional spaces. The embedding imposes a well-structured
hypercube topology on the search space, endows it with a semantic-aware
neighbourhood, and enables convenient search using Cartesian coordinates.
The embedding algorithm consists in locality-driven optimization
and operates in abstraction from a specific fitness function, improving
locality of all possible fitness landscapes simultaneously. In experimental
part, we validate the approach on the domain of symbolic regression
and demonstrate on a large sample of problem instances that semantic
embedding provides better search performance than the original program
space. Moreover, we show also that semantic embedding of small programs
can be effectively exploited in a compositional manner to search
the space of large compound programs.
BibTeX:
@inproceedings{DBLP:conf/gecco/Krawiec11,
  author = {Krzysztof Krawiec},
  title = {Semantically embedded genetic programming: automated design of abstract program representations},
  booktitle = {Proceedings of the 13th Annual Conference on Genetic and Evolutionary Computation},
  publisher = {ACM},
  year = {2011},
  pages = {1379--1386},
  url = {http://doi.acm.org/10.1145/2001576.2001762},
  doi = {http://doi.org/10.1145/2001576.2001762}
}
Krawiec, K. Learnable Embeddings of Program Spaces 2011
Vol. 6621Proceedings of the 14th European Conference on Genetic Programming, EuroGP 2011, pp. 167-178 
inproceedings  
Abstract: We consider a class of adaptive, globally-operating, semantic-based
embeddings of programs into discrete multidimensional spaces termed
prespaces. In the proposed formulation, the original space of programs
and its prespace are bound with a learnable mapping, where the process
of learning is aimed at improving the overall locality of the new
representation with respect to program semantics. To learn the mapping,
which is formally a permutation of program locations in the prespace,
we propose two algorithms: simple greedy heuristics and an evolutionary
algorithm. To guide the learning process, we use a new definition
of semantic locality. In an experimental illustration concerning
four symbolic regression domains, we demonstrate that an evolutionary
algorithm is able to improve the embedding designed by means of greedy
search, and that the learnt prespaces usually offer better search
performance than the original program space.
BibTeX:
@inproceedings{krawiec:2011:EuroGP,
  author = {Krzysztof Krawiec},
  title = {Learnable Embeddings of Program Spaces},
  booktitle = {Proceedings of the 14th European Conference on Genetic Programming, EuroGP 2011},
  publisher = {Springer Verlag},
  year = {2011},
  volume = {6621},
  pages = {167--178}
}
Krawiec, K., Jaśkowski, W. and Szubert, M. Evolving Small-Board Go Players using Coevolutionary Temporal Difference Learning with Archive 2011 International Journal of Applied Mathematics and Computer Science
Vol. 21(4), pp. 717-731 
article DOI URL 
Abstract: We apply Coevolutionary Temporal Difference Learning (CTDL) to learn
small-board Go strategies represented as weighted piece counters.
CTDL is a randomized learning technique which interleaves two search
processes that operate in intra-game and inter-game mode. The intra-game
learning is driven by gradient-descent Temporal Difference Learning
(TDL), a reinforcement learning method that updates the board evaluation
function according to differences observed between its values for
consecutively visited game states. For the inter-game learning component,
we provide coevolutionary algorithm that maintains a sample of strategies
and uses the outcomes of games played between them to iteratively
modify the probability distribution, according to which new strategies
are generated and added to the sample. We analyze CTDL’s
sensitivity to all important parameters, including the trace decay
constant that controls the lookahead horizon of TDL, and the relative
intensity of intra-game and inter-game learning. We investigate also
how the presence of memory (an archive) affects the search performance,
and find out that the archived approach is superior to other techniques
considered here, and produces strategies that outperform a handcrafted
weighted piece counter strategy and a simple liberty-based heuristics.
This encouraging result can be potentially generalized not only to
other strategy representations used for small-board Go, but also
to different games and a broader class of problems, because CTDL
is generic and does not rely on any problem-specific knowledge.
BibTeX:
@article{2011AMCSKrawiecJaskowskiSzubert,
  author = {Krzysztof Krawiec and Wojciech Jaśkowski and Marcin Szubert},
  title = {Evolving Small-Board Go Players using Coevolutionary Temporal Difference Learning with Archive},
  journal = {International Journal of Applied Mathematics and Computer Science},
  year = {2011},
  volume = {21},
  number = {4},
  pages = {717--731},
  url = {http://www.cs.put.poznan.pl/kkrawiec/pubs/2011AMCSpreprint.pdf},
  doi = {DOI: 10.2478/v10006-011-0057-3}
}
Krawiec, K. and Nawrocki, M. Evolutionary Tuning of Compound Image Analysis Systems for Effective Li cense Plate Recognition 2011
Vol. 6922Computational Collective Intelligence. Technologies and Application s, pp. 203-212 
inproceedings URL 
BibTeX:
@inproceedings{2011ICCCINawrockiKrawiec,
  author = {Krawiec, Krzysztof and Nawrocki, Mateusz},
  title = {Evolutionary Tuning of Compound Image Analysis Systems for Effective Li cense Plate Recognition},
  booktitle = {Computational Collective Intelligence. Technologies and Application s},
  publisher = {Springer Berlin / Heidelberg},
  year = {2011},
  volume = {6922},
  pages = {203-212},
  url = {http://dx.doi.org/10.1007/978-3-642-23935-9_20}
}
Moraglio, A., Krawiec, K. and Johnson, C. Geometric Semantic Genetic Programming 2011 5th workshop on Theory of Randomized Search Heuristics  inproceedings URL 
Abstract: Traditional Genetic Programming (GP) searches the space of functions/programs
by using search operators that manipulate their syntactic representation
(e.g., parse trees), regardless of their semantic. Recently, semantically
aware search operators have been shown to outperform purely syntactic
operators. In this work, using a formal geometric view on search
operators and representations, we bring the semantic approach to
its extreme consequences and introduce a novel form of GP - Semantic
GP (SGP) - that searches directly the space of the underlying semantic
of the programs. This perspective provides new insights on the relation
between syntax, semantic, search operators and fitness landscape,
and allows for the principled formal design of semantic search operators
for different classes of problems. We derive specific forms of SGP
for a number of classic GP domains and report preliminary experimental
results. Furthermore, we show that the search of SGP is equivalent
to the search of a traditional Genetic Algorithm on the One-Max problem.
BibTeX:
@inproceedings{2011ThrashMoraglioKrawiecJohnson,
  author = {Alberto Moraglio and Krzysztof Krawiec and Colin Johnson},
  title = {Geometric Semantic Genetic Programming},
  booktitle = {5th workshop on Theory of Randomized Search Heuristics},
  year = {2011},
  url = {http://www.thrash-workshop.org/slides/moraglio.pdf}
}
Szubert, M., Jaśkowski, W. and Krawiec, K. Learning Board Evaluation Function for Othello by Hybridizing Coevolution with Temporal Difference Learning 2011 Control and Cybernetics
Vol. 40(3), pp. 805-832 
article  
Abstract: Hybridization of global and local search techniques has already produced
promising results in the fields of optimization and machine learning.
It is commonly presumed that approaches employing this idea, like
memetic algorithms that combine evolutionary algorithms with local
search, benefit from complementary characteristics of constituent
methods and maintain the right balance between exploration and exploitation
of the search space. While such extensions of evolutionary algorithms
have been intensively studied, hybrids of local search with coevolutionary
algorithms have not received much attention yet. In this paper we
attempt to fill this gap by presenting Coevolutionary Temporal Difference
Learning (CTDL) that works by interlacing global search provided
by competitive coevolution and local search by means of temporal
difference learning. We verify CTDL by applying it to the board game
of Othello, where it learns board evaluation functions represented
by a linear architecture of weighted piece counter. The results of
a computational experiment show CTDL’s superiority when compared
to coevolutionary algorithm and temporal difference learning alone,
both in terms of performance of elaborated strategies and computational
cost. In order to further exploit CTDL’s potential, we also extend
it by an archive that keeps track of selected well-performing solutions
found so far and uses them to improve search convergence. The overall
conclusion is that the fusion of various forms of coevolution with
a gradient-based local search can be highly beneficial and deserves
further research.
BibTeX:
@article{szubert11learning,
  author = {Marcin Szubert and Wojciech Jaśkowski and Krzysztof Krawiec},
  title = {Learning Board Evaluation Function for Othello by Hybridizing Coevolution with Temporal Difference Learning},
  journal = {Control and Cybernetics},
  year = {2011},
  volume = {40},
  number = {3},
  pages = {805--832}
}
Jaśkowski, W. and Krawiec, K. Coordinate System Archive for coevolution 2010 IEEE Congress on Evolutionary Computation (CEC 2010), pp. 1-10  inproceedings DOI  
Abstract: Problems in which some entities interact with each other are common in computational intelligence. This scenario, typical for co-evolving artificial-life agents, learning strategies for games, and machine learning from examples, can be formalized as test-based problem. In test-based problems, candidate solutions are evaluated on a number of test cases (agents, opponents, examples). It has been recently shown that at least some of such problems posses underlying problem structure, which can be formalized in a notion of coordinate system, which spatially arranges candidate solutions and tests in a multidimensional space. Such a coordinate system can be extracted to reveal underlying objectives of the problem, which can be then further exploited to help coevolutionary algorithm make progress. In this study, we propose a novel coevolutionary archive method, called Coordinate System Archive (COSA) that is based on these concepts. In the experimental part, we compare COSA to two state-of-the-art archive methods, IPCA and LAPCA. Using two different objective performance measures, we find out that COSA is superior to these methods on a class of artificial problems (COMPARE-ON-ONE).
BibTeX:
@inproceedings{DBLP:conf/cec/JaskowskiK10,
  author = {Wojciech Jaśkowski and Krzysztof Krawiec},
  title = {Coordinate System Archive for coevolution},
  booktitle = {IEEE Congress on Evolutionary Computation (CEC 2010)},
  publisher = {IEEE Press},
  year = {2010},
  pages = {1-10},
  doi = {http://doi.org/10.1109/CEC.2010.5586066}
}
Krawiec, K., Kukawka, B. and Maciejewski, T. Evolving cascades of voting feature detectors for vehicle detection in satellite imagery 2010 IEEE Congress on Evolutionary Computation (CEC 2010), pp. 2392-2399  inproceedings DOI  
Abstract: We propose an evolutionary method for detection of vehicles in satellite
imagery which involves a large number of simple elementary features
and multiple detectors trained by genetic programming. The complete
detection system is composed of several detectors that are chained
into a cascade and successively filter out the negative examples.
Each detector is a committee of genetic programming trees that together
vote over the decision concerning vehicle presence, and is trained
only on the examples classified as positive by the previous cascade
node. The individual trees use typical arithmetic transformations
to aggregate features selected from a very large collections of Haar-like
features derived from the input image. The paper presents detailed
description of the proposed algorithm and reports the results of
an extensive computational experiment carried out on real-world satellite
images. The evolved detection system exhibits competitive sensitivity
and relatively low false positive rate for testing images, despite
not making use of domain-specific knowledge.
BibTeX:
@inproceedings{2010CECKrawiecKukawkaMaciejewski,
  author = {Krzysztof Krawiec and Bartosz Kukawka and Tomasz Maciejewski},
  title = {Evolving cascades of voting feature detectors for vehicle detection in satellite imagery},
  booktitle = {IEEE Congress on Evolutionary Computation (CEC 2010)},
  publisher = {IEEE Press},
  year = {2010},
  pages = {2392--2399},
  doi = {http://doi.org/10.1109/CEC.2010.5586155}
}
Krawiec, K. and Lichocki, P. Using Co-solvability to Model and Exploit Synergetic Effects in Evolution 2010
Vol. 6239Parallel Problem Solving from Nature -- PPSN XI, pp. 492-501 
inproceedings  
Abstract: We introduce, analyze, and experimentally examine co-solvability,
an ability of a solution to solve a pair of fitness cases (tests).
Based on this concept, we devise a co-solvability fitness function
that makes solutions compete for rewards granted for solving pairs
of tests, in a way analogous to implicit fitness sharing. We prove
that co-solvability fitness function is by definition synergistic
and imposes selection pressure which is qualitatively different from
that of standard fitness function or implicit fitness sharing. The
results of experimental verification on eight genetic programming
tasks demonstrate that evolutionary runs driven by co-solvability
fitness function usually converge faster to well-performing solutions
and are more likely to reach global optima.
BibTeX:
@inproceedings{LNCS62390492,
  author = {Krzysztof Krawiec and Paweł Lichocki},
  title = {Using Co-solvability to Model and Exploit Synergetic Effects in Evolution},
  booktitle = {Parallel Problem Solving from Nature -- PPSN XI},
  publisher = {Springer},
  year = {2010},
  volume = {6239},
  pages = {492--501}
}
Krawiec, K. and Szubert, M. Coevolutionary Temporal Difference Learning for small-board Go 2010 Evolutionary Computation (CEC), 2010 IEEE Congress on, pp. 1-8  inproceedings DOI  
Abstract: In this paper we apply Coevolutionary Temporal Difference Learning (CTDL), a hybrid of coevolutionary search and reinforcement learning proposed in our former study, to evolve strategies for playing the game of Go on small boards (5×5). CTDL works by interlacing exploration of the search space provided by one-population competitive coevolution and exploitation by means of temporal difference learning. Despite using simple representation of strategies (weighted piece counter), CTDL proves able to evolve players that defeat solutions found by its constituent methods. The results of the conducted experiments indicate that our algorithm turns out to be superior to pure coevolution and pure temporal difference learning, both in terms of performance of the elaborated strategies and the computational cost. This demonstrates the existence of synergistic interplay between components of CTDL, which we also briefly discuss in this study.
BibTeX:
@inproceedings{2010CECKrawiecSzubert,
  author = {Krzysztof Krawiec and Marcin Szubert},
  title = {Coevolutionary Temporal Difference Learning for small-board Go},
  booktitle = {Evolutionary Computation (CEC), 2010 IEEE Congress on},
  year = {2010},
  pages = {1-8},
  doi = {http://doi.org/10.1109/CEC.2010.5586054}
}
Krawiec, K. and Wieloch, B. Automatic generation and exploitation of related problems in genetic programming 2010 IEEE Congress on Evolutionary Computation (CEC 2010), pp. 1248-1255  inproceedings DOI  
Abstract: We propose an evolutionary framework that uses the set of instructions
provided with a genetic programming (GP) problem to automatically
build a repertoire of related problems and subsequently uses them
to improve the performance of search. The novel idea is to use the
synthesised related problems to simultaneously exert multiple selection
pressures on the evolving population(s). For that framework, we design
two methods. In the first method, individuals optimising for particular
problems dwell in separate populations and spawn clones which migrate
to other populations, similarly to the island model. The second method
operates on a single population and ranks the fitness values that
individuals receive from particular problems to make them comparable.
When applied to six symbolic regression problems of different difficulty,
both methods perform better than the standard GP, though sometimes
fail to prove superior to certain control setup.
BibTeX:
@inproceedings{2010CECKrawiecWieloch,
  author = {Krzysztof Krawiec and Bartosz Wieloch},
  title = {Automatic generation and exploitation of related problems in genetic programming},
  booktitle = {IEEE Congress on Evolutionary Computation (CEC 2010)},
  publisher = {IEEE Press},
  year = {2010},
  pages = {1248--1255},
  doi = {http://doi.org/10.1109/CEC.2010.5586120}
}
Nawrocki, M. and Krawiec, K. A Robuts Method for Real-Time Detection and Recognition of Licence Plates 2010 IEEE International Conference on Multimedia Communications, Services, and Security (MCSS 2010), pp. 171-175  inproceedings  
BibTeX:
@inproceedings{2010MCSSNawrockiKrawiec,
  author = {Mateusz Nawrocki and Krzysztof Krawiec},
  title = {A Robuts Method for Real-Time Detection and Recognition of Licence Plates},
  booktitle = {IEEE International Conference on Multimedia Communications, Services, and Security (MCSS 2010)},
  year = {2010},
  pages = {171--175}
}
Jaśkowski, W. and Krawiec, K. Formal Analysis and Algorithms for Extracting Coordinate Systems of Games 2009 IEEE Symposium on Computational Intelligence and Games, pp. 201-208  inproceedings URL 
BibTeX:
@inproceedings{2009CIGBucci,
  author = {Wojciech Jaśkowski and Krzysztof Krawiec},
  title = {Formal Analysis and Algorithms for Extracting Coordinate Systems of Games},
  booktitle = {IEEE Symposium on Computational Intelligence and Games},
  publisher = { IEEE },
  year = {2009},
  pages = {201--208},
  url = {http://www.ieee-cig.org/cig-2009/Proceedings/proceedings/papers/cig2009_028e.pdf}
}
Jaśkowski, W., Krawiec, K. and Wieloch, B. Genetic Programming for Generative Learning and Recognition of Hand-drawn Shapes 2009
Vol. 213Evolutionary Image Analysis and Signal Processing, pp. 73-90 
inbook URL 
BibTeX:
@inbook{2007SpringerBook,
  author = {Wojciech Jaśkowski and Krzysztof Krawiec and Bartosz Wieloch},
  title = {Genetic Programming for Generative Learning and Recognition of Hand-drawn Shapes},
  booktitle = {Evolutionary Image Analysis and Signal Processing},
  publisher = {Springer-Verlag},
  year = {2009},
  volume = {213},
  pages = {73--90},
  url = {http://www.springer.com/engineering/book/978-3-642-01635-6}
}
Krawiec, K. and Lichocki, P. Approximating geometric crossover in semantic space 2009 GECCO '09: Proceedings of the 11th Annual conference on Genetic and evolutionary computation, pp. 987-994  inproceedings URL 
Abstract: We propose a crossover operator that works with genetic programming
trees and is approximately geometric crossover in the semantic space.
By defining semantic as program's evaluation profile with respect
to a set of fitness cases and constraining to a specific class of
metric-based fitness functions, we cause the fitness landscape in
the semantic space to have perfect fitness-distance correlation.
The proposed approximately geometric semantic crossover exploits
this property of the semantic fitness landscape by an appropriate
sampling. We demonstrate also how the proposed method may be conveniently
combined with hill climbing. We discuss the properties of the methods,
and describe an extensive computational experiment concerning logical
function synthesis and symbolic regression.
BibTeX:
@inproceedings{2009GECCOSemanticXover,
  author = {Krzysztof Krawiec and Pawel Lichocki},
  title = {Approximating geometric crossover in semantic space},
  booktitle = {GECCO '09: Proceedings of the 11th Annual conference on Genetic and evolutionary computation},
  publisher = {ACM},
  year = {2009},
  pages = {987--994},
  url = {http://doi.acm.org/10.1145/1569901.1570036}
}
Krawiec, K. and Wieloch, B. Functional modularity for genetic programming 2009 GECCO 09: Proceedings of the 11th Annual conference on Genetic and evolutionary computation, pp. 995-1002  inproceedings URL 
Abstract: In this paper we introduce, formalize, and experimentally validate
a novel concept of functional modularity for Genetic Programming
(GP). We rely on module definition that is most natural for GP: a
piece of program code (subtree). However, as opposed to syntax-based
approaches that abstract from the actual computation performed by
a module, we analyze also its semantic using a set of fitness cases.
In particular, the central notion of this approach is subgoal, an
entity that embodies module's desired semantic and is used to evaluate
module candidates. As the cardinality of the space of all subgoals
is exponential with respect to the number of fitness cases, we introduce
monotonicity to assess subgoals' potential utility for searching
for good modules. For a given subgoal and a sample of modules, monotonicity
measures the correlation of subgoal's distance from module's semantics
and the fitness of the solution the module is part of. In the experimental
part we demonstrate how these concepts may be used to describe and
quantify the modularity of two simple problems of Boolean function
synthesis. In particular, we conclude that monotonicity usefully
differentiates two problems with different nature of modularity,
allows us to tell apart the useful subgoals from the other ones,
and may be potentially used for problem decomposition and enhance
the efficiency of evolutionary search.
BibTeX:
@inproceedings{2009GECCOModules,
  author = {Krzysztof Krawiec and Bartosz Wieloch},
  title = {Functional modularity for genetic programming},
  booktitle = {GECCO 09: Proceedings of the 11th Annual conference on Genetic and evolutionary computation},
  publisher = {ACM},
  year = {2009},
  pages = {995--1002},
  url = {http://doi.acm.org/10.1145/1569901.1570037}
}
Krawiec, K. and Wieloch, B. Analysis of Semantic Modularity for Genetic Programming 2009 Foundations of Computing and Decision Sciences
Vol. 34(4), pp. 265-285 
article  
Abstract: In this paper we analyze the properties of functional modularity,
a concept introduced in [DBLP:conf/gecco/KrawiecW09] for detecting
and measuring modularity in problems of automatic program synthesis,
in particular by means of genetic programming. The basic components
of functional modularity approach are subgoals -- entities that embody
module's semantic -- and monotonicity, a measure for assessing subgoals'
potential utility for searching for good modules. For a given subgoal
and a sample of solutions decomposed into parts and contexts according
to module definition, monotonicity measures the correlation of distance
between semantics of solution's part and the fitness of the solution.
The central tenet of this approach is that highly monotonous subgoals
can be used to decompose the task and improve search convergence.
In the experimental part we investigate the properties of functional
modularity using eight instances of problems of Boolean function
synthesis. The results show that monotonicity varies depending on
problem's structure of modularity and correctly identifies good subgoals,
potentially enabling automatic program decomposition.
BibTeX:
@article{2009KrawiecWielochFCDS,
  author = {Krzysztof Krawiec and Bartosz Wieloch},
  title = {Analysis of Semantic Modularity for Genetic Programming},
  journal = {Foundations of Computing and Decision Sciences},
  year = {2009},
  volume = {34},
  number = {4},
  pages = {265--285}
}
Lichocki, P., Krawiec, K. and Jaśkowski, W. Evolving Teams of Cooperating Agents for Real-Time Strategy Game 2009 Applications of Evolutionary Computing, pp. 333-342  inproceedings  
Abstract: We apply gene expression programing to evolve a player for a real-time
strategy (RTS) video game. The paper describes the game, evolutionary
encoding of strategies and the technical implementation of experimental
framework. In the experimental part, we compare two setups that differ
with respect to the used approach of task decomposition. One of the
setups turns out to be able to evolve an effective strategy, while
the other leads to more sophisticated yet inferior solutions. We
discuss both the quantitative results and the behavioral patterns
observed in the evolved strategies.
BibTeX:
@inproceedings{2009LichockiKrawiecEvoGames,
  author = {Paweł Lichocki and Krzysztof Krawiec and Wojciech Jaśkowski},
  title = {Evolving Teams of Cooperating Agents for Real-Time Strategy Game},
  booktitle = {Applications of Evolutionary Computing},
  publisher = {Springer},
  year = {2009},
  pages = {333--342}
}
Perzyk, M., Krawiec, K. and Kozłowski, J. Application of time-series analysis in foundry production 2009 Archives of Foundry Engineering
Vol. 9(3), pp. 109-114 
article  
Abstract: Characterization of the time-series analysis is presented, as a data
mining tool which facilitates better understanding nature of manufacturing
process and permits forecasting of future values of the process parameters
or production results on the basis of the past data, recorded in
regular intervals. The main methods and problems of the time-series
analysis are presented, related to the trend function, evaluation
of seasonality and significance of the information contents in the
residual values. The authors’ research results, related to exemplary
production data collected in a foundry with Disamatic molding line
(temperature of the molding sand), are presented. It is concluded
that a properly performed analysis of time-series can be a useful
tool for analysis and predictions of foundry production process.
BibTeX:
@article{KrawiecPerzykKozlowski2009,
  author = {Marcin Perzyk and Krzysztof Krawiec and Jacek Kozłowski},
  title = {Application of time-series analysis in foundry production},
  journal = {Archives of Foundry Engineering},
  year = {2009},
  volume = {9},
  number = {3},
  pages = {109--114}
}
Szubert, M., Jaśkowski, W. and Krawiec, K. Coevolutionary Temporal Difference Learning for Othello 2009 IEEE Symposium on Computational Intelligence and Games, pp. 104-111  inproceedings URL 
BibTeX:
@inproceedings{2009CIGOthello,
  author = {Marcin Szubert and Wojciech Jaśkowski and Krzysztof Krawiec},
  title = {Coevolutionary Temporal Difference Learning for Othello},
  booktitle = {IEEE Symposium on Computational Intelligence and Games},
  publisher = { IEEE },
  year = {2009},
  pages = {104-111},
  url = {http://www.ieee-cig.org/cig-2009/Proceedings/proceedings/papers/cig2009_015e.pdf}
}
Gajda, P. and Krawiec, K. Evolving a Vision-Driven Robot Controller for Real-World Indoor Navigation 2008
Vol. 4974Applications of Evolutionary Computing, pp. 184-193 
inproceedings  
Abstract: In this paper, we use genetic programming (GP) to evolve a vision-driven
robot controller capable of navigating in a real-world environment.
To this aim, we extract visual primitives from the video stream provided
by a camera mounted on the robot and let them to be interpreted by
a GP individual. The response of GP expressions is then used to control
robot's servos. Thanks to the primitive-based approach, evolutionary
process is less constrained in the process of synthesizing image
features. Experiments concerning navigation in indoor environment
indicate that the evolved controller performs quite well despite
very limited human intervention in the design phase.
BibTeX:
@inproceedings{krawiec08robot,
  author = {Paweł Gajda and Krzysztof Krawiec},
  title = {Evolving a Vision-Driven Robot Controller for Real-World Indoor Navigation},
  booktitle = {Applications of Evolutionary Computing},
  publisher = {Springer},
  year = {2008},
  volume = {4974},
  pages = {184--193},
  note = {LNCS49740184}
}
Jaśkowski, W., Krawiec, K. and Wieloch, B. Winning Ant Wars: Evolving a Human-Competitive Game Strategy Using Fitnessless Selection 2008
Vol. 4971Genetic Programming, pp. 13-24 
inproceedings  
Abstract: This paper tells the story of BrilliAnt, the winner of the Ant Wars
contest organized within GECCO'2007, Genetic and Evolutionary Computation
Conference. The task for the Ant Wars contestants was to evolve a
controller for a virtual ant that collects food in a square toroidal
grid environment in the presence of a competing ant. BrilliAnt, submitted
to the contest by our team, has been evolved through competitive
one-population coevolution using genetic programming and a novel
fitnessless selection method. In the paper, we detail the evolutionary
setup that lead to BrilliAnt's emergence, assess its human-competitiveness,
and describe selected behavioral patterns observed in its strategy.
BibTeX:
@inproceedings{jaskowski08antwars,
  author = {Wojciech Jaśkowski and Krzysztof Krawiec and Bartosz Wieloch},
  title = {Winning Ant Wars: Evolving a Human-Competitive Game Strategy Using Fitnessless Selection},
  booktitle = {Genetic Programming},
  publisher = {Springer},
  year = {2008},
  volume = {4971},
  pages = {13--24},
  note = {LNCS49710013}
}
Jaśkowski, W., Krawiec, K. and Wieloch, B. NeuroHunter - An Entry for the Balanced Diet Contest 2008 (RA-010/08)School: Poznan University of Technology, Institute of Computing Science  techreport URL 
Abstract: NeuroHunter is the name of an evolved neural network that won the
GECCO'2008 Balanced Diet contest, organized within GECCO'2008, Genetic
and Evolutionary Computation Conference, in Atlanta, Georgia. The
goal was to evolve a controller for a virtual agent that collects
two types of food in a 256x256 board, trying to maximize the amount
of food collected while minimizing the imbalance between the two
types of food. The major difficulty of the task consists in partial
observability: the agent cannot see the food pieces. However, the
probability of finding a food piece of particular type at a particular
location depends on another feature, elevation, which is visible
to the agent. Thus, the agent has to learn this dependency, also
avoiding revisiting the same locations. Our winner entry used the
HyperNEAT by Stanley et al., an evolved structural neural net. The
report describes in detail the method, technology, and the results.
BibTeX:
@techreport{Jaskowski08BalancedDiet,
  author = {Wojciech Jaśkowski and Krzysztof Krawiec and Bartosz Wieloch},
  title = {NeuroHunter - An Entry for the Balanced Diet Contest},
  school = {Poznan University of Technology, Institute of Computing Science},
  year = {2008},
  number = {RA-010/08},
  url = {http://www.cs.put.poznan.pl/kkrawiec/pubs/2008GECCOBalancedDiet.pdf}
}
Jaśkowski, W., Krawiec, K. and Wieloch, B. Multitask Visual Learning using Genetic Programming 2008 Evolutionary Computation
Vol. 16(4), pp. 439-459 
article  
Abstract: We propose a multitask learning method of visual concepts within the
genetic program- ming (GP) framework. Each GP individual is composed
of several trees that process visual primitives derived from input
images. Two trees solve two different visual tasks and are allowed
to share knowledge with each other by commonly calling the remain-
ing GP trees (subfunctions) included in the same individual. The
performance of a particular tree is measured by its ability to reproduce
the shapes contained in the train- ing images. We apply this method
to visual learning tasks of recognizing simple shapes and compare
it to a reference method. The experimental verification demonstrates
that such multitask learning often leads to performance improvements
in one or both solved tasks, without extra computational effort.
BibTeX:
@article{jaskowski08ec,
  author = {Wojciech Jaśkowski and Krzysztof Krawiec and Bartosz Wieloch},
  title = {Multitask Visual Learning using Genetic Programming},
  journal = {Evolutionary Computation},
  publisher = {MIT Press},
  year = {2008},
  volume = {16},
  number = {4},
  pages = {439--459}
}
Jaśkowski, W., Krawiec, K. and Wieloch, B. Multi-task code reuse in genetic programming 2008 GECCO '08: Proceedings of the 2008 GECCO conference companion on Genetic and evolutionary computation, pp. 2159-2164  inproceedings DOI  
Abstract: We propose a method of knowledge reuse between evolutionary processes
that solve different optimization tasks. We define the method in
the framework of tree-based genetic programming (GP) and implement
it as code reuse between GP trees that evolve in parallel in separate
populations delegated to particular tasks. The technical means of
code reuse is a crossbreeding operator which works very similar to
standard tree-swapping crossover. We consider two variants of this
operator, which differ in the way they handle the incompatibility
of terminals between the considered problems. In the experimental
part we demonstrate that such code reuse is usually beneficial and
leads to success rate improvements when solving the common boolean
benchmarks.
BibTeX:
@inproceedings{jaskowski08reuse,
  author = {Wojciech Jaśkowski and Krzysztof Krawiec and Bartosz Wieloch},
  title = {Multi-task code reuse in genetic programming},
  booktitle = {GECCO '08: Proceedings of the 2008 GECCO conference companion on Genetic and evolutionary computation},
  publisher = {ACM},
  year = {2008},
  pages = {2159--2164},
  doi = {http://doi.org/10.1145/1388969.1389040}
}
Jaśkowski, W., Krawiec, K. and Wieloch, B. Fitnessless coevolution 2008 GECCO '08: Proceedings of the 10th annual conference on Genetic and evolutionary computation, pp. 355-362  inproceedings DOI  
BibTeX:
@inproceedings{jaskowski08fitnessless,
  author = {Jaśkowski, Wojciech and Krawiec, Krzysztof and Wieloch, Bartosz},
  title = {Fitnessless coevolution},
  booktitle = {GECCO '08: Proceedings of the 10th annual conference on Genetic and evolutionary computation},
  publisher = {ACM},
  year = {2008},
  pages = {355--362},
  doi = {http://doi.org/10.1145/1389095.1389161}
}
Jaśkowski, W., Krawiec, K. and Wieloch, B. Evolving Strategy for a Probabilistic Game of Imperfect Information using Genetic Programming 2008 Genetic Programming and Evolvable Machines
Vol. 9(4), pp. 281-294 
article DOI  
Abstract: We provide the complete record of methodology that let us evolve BrilliAnt,
the winner of the Ant Wars contest. Ant Wars contestants are virtual
ants collecting food on a grid board in the presence of a competing
ant. BrilliAnt has been evolved through a competitive one-population
coevolution using genetic programming and fitnessless selection.
In this paper, we detail the evolutionary setup that lead to BrilliAnt's
emergence, assess its direct and indirect human-competitiveness,
and describe the behavioral patterns observed in its strategy.
BibTeX:
@article{jaskowski08gpem,
  author = {Wojciech Jaśkowski and Krzysztof Krawiec and Bartosz Wieloch},
  title = {Evolving Strategy for a Probabilistic Game of Imperfect Information using Genetic Programming},
  journal = {Genetic Programming and Evolvable Machines},
  year = {2008},
  volume = {9},
  number = {4},
  pages = {281--294},
  doi = {http://doi.org/10.1007/s10710-008-9062-1}
}
Krawiec, K. and Polewski, P. Potential fitness for genetic programming 2008 GECCO '08: Proceedings of the 2008 GECCO conference companion on Genetic and evolutionary computation, pp. 2175-2180  inproceedings DOI  
Abstract: We introduce potential fitness, a variant of fitness function that
operates in the space of schemata and is applicable to tree-based
genetic programing. The proposed evaluation al- gorithm estimates
the maximum possible gain in fitness of an individual's direct offspring.
The value of the potential fitness is calculated by analyzing the
context semantics and subtree semantics for all contexts (schemata)
of the evalu- ated tree. The key feature of the proposed approach
is that a tree is rewarded for the correctly classiffed fitness cases,
but it is not penalized for the incorrectly classified ones, provided
that such errors are recoverable by substitution of an appropriate
subtree (which is however not explicitly con- sidered by the algorithm).
The experimental evaluation on a set of seven boolean benchmarks
shows that the use of potential fitness may lead to better convergence
and higher success rate of the evolutionary run.
BibTeX:
@inproceedings{krawiec08potential,
  author = {Krzysztof Krawiec and Przemysław Polewski},
  title = {Potential fitness for genetic programming},
  booktitle = {GECCO '08: Proceedings of the 2008 GECCO conference companion on Genetic and evolutionary computation},
  publisher = {ACM},
  year = {2008},
  pages = {2175--2180},
  doi = {http://doi.org/10.1145/1388969.1389043}
}
Ignaszak, Z., Popielarski, P. and Krawiec, K. Contribution to quantitative identification of casting defects based on computer analysis of X-ray images 2007 Archives of Foundry Engineering
Vol. 7(4), pp. 89-94 
article  
Abstract: The forecast of structure and properties of casting is based on results
of computer simulation of physical processes which are carried out
during the casting processes. For the effective using of simulation
system it is necessary to validate mathematica-physical models describing
process of casting formation and the creation of local discontinues,
witch determinate the casting properties. In the paper the proposition
for quantitative validation of VP system using solidification casting
defects by information sources of II group (methods of NDT) was introduced.
It was named the VP/RT validation (virtual prototyping/radiographic
testing validation). Nowadays identification of casting defects noticeable
on X-ray images bases on comparison of X-ray image of casting with
relates to the ASTM. The results of this comparison are often not
conclusive because based on operator’s subjective assessment. In
the paper the system of quantitative identification of iron casting
defects on X-ray images and classification this defects to ASTM class
is presented. The methods of pattern recognition and machine learning
were applied.
BibTeX:
@article{IgnaszakPopielarskiKrawiec07,
  author = {Zenon Ignaszak and Paweł Popielarski and Krzysztof Krawiec},
  title = {Contribution to quantitative identification of casting defects based on computer analysis of X-ray images},
  journal = {Archives of Foundry Engineering},
  year = {2007},
  volume = {7},
  number = {4},
  pages = {89-94}
}
Jaśkowski, W., Krawiec, K. and Wieloch, B. Learning and Recognition of Hand-Drawn Shapes Using Generative Genetic Programming 2007
Vol. 4448EvoWorkshops 2007, pp. 281-290 
inproceedings  
Abstract: We describe a novel method of evolutionary visual learning that uses
generative approach for assessing learner’s ability to recognize
image contents. Each learner, implemented as a genetic programming
individual, processes visualprimitivesthatrepresentlocalsalientfeatures
derived from a raw input raster image. In response to that input,
the learnerproducespartialreproductionoftheinputimage,andisevaluated
according to the quality of that reproduction. We present the method
in detail and verify it experimentally on the real-world task of
recognition of hand-drawn shapes.
BibTeX:
@inproceedings{jaskowski07learning,
  author = {Wojciech Jaśkowski and Krzysztof Krawiec and Bartosz Wieloch},
  title = {Learning and Recognition of Hand-Drawn Shapes Using Generative Genetic Programming},
  booktitle = {EvoWorkshops 2007},
  publisher = {Springer-Verlag},
  year = {2007},
  volume = {4448},
  pages = {281-290}
}
Jaśkowski, W., Krawiec, K. and Wieloch, B. Knowledge Reuse in Genetic Programming applied to Visual Learning 2007 Genetic and Evolutionary Computation Conference GECCO, pp. 1790-1797  inproceedings URL 
Abstract: We propose a method of knowledge reuse for an ensemble of genetic
programming-based learners solving a visual learning task. First,
we introduce a visual learning method that uses genetic programming
individuals to represent hypotheses. Individuals-hypotheses process
image representation composed of visual primitives derived from the
training images that contain objects to be recognized. The process
of recognition is generative, i.e., an individual is supposed to
restore the shape of the processed object by drawing its reproduction
on a separate canvas. This canonical method is extended with a knowledge
reuse mechanism that allows a learner to import genetic material
from hypotheses that evolved for the other decision classes (object
classes). We compare the performance of the extended approach to
the basic method on a real-world tasks of handwritten character recognition,
and conclude that knowledge reuse leads to signifcant convergence
speedup and, more importantly, significantly reduces the risk of
overfitting.
BibTeX:
@inproceedings{Jaskowski07geccoGBML,
  author = {Wojciech Jaśkowski and Krzysztof Krawiec and Bartosz Wieloch},
  title = {Knowledge Reuse in Genetic Programming applied to Visual Learning},
  booktitle = {Genetic and Evolutionary Computation Conference GECCO},
  publisher = {Association for Computing Machinery},
  year = {2007},
  pages = {1790--1797},
  url = {http://www.cs.put.poznan.pl/kkrawiec/pubs/p1790.pdf}
}
Jaśkowski, W., Krawiec, K. and Wieloch, B. Knowledge Reuse for an Ensemble of GP-based Learners 2007
Vol. 160Evolutionary Computation and Global Optimization 2007, pp. 135 - 142 
inproceedings URL 
BibTeX:
@inproceedings{Jaskowski07kaeiog,
  author = {Wojciech Jaśkowski and Krzysztof Krawiec and Bartosz Wieloch},
  title = {Knowledge Reuse for an Ensemble of GP-based Learners},
  booktitle = {Evolutionary Computation and Global Optimization 2007},
  publisher = {Oficyna Wydawnicza Politechniki Warszawskiej},
  year = {2007},
  volume = {160},
  pages = {135 -- 142},
  url = {http://www.cs.put.poznan.pl/kkrawiec/pubs/2007KAEiOG.pdf}
}
Jaśkowski, W., Krawiec, K. and Wieloch, B. Genetic Programming for Cross-task Knowledge Sharing 2007 Genetic and Evolutionary Computation Conference GECCO, pp. 1620-1627  inproceedings URL 
Abstract: We consider multitask learning of visual concepts within genetic programming
(GP) framework. The proposed method evolves a population of GP individuals,
with each of them composed of several GP trees that process visual
primitives derived from input images. The two main trees are delegated
to solving two different visual tasks and are allowed to share knowledge
with each other by calling the remaining GP trees (subfunctions)
included in the same individual. The method is applied to the visual
learning task of recognizing simple shapes, using generative approach
based on visual primitives. We compare this approach to a reference
method devoid of knowledge sharing, and conclude that in the worst
case cross-task learning performs equally well, and in many cases
it leads to significant performance improvements in one or both solved
tasks.
BibTeX:
@inproceedings{Jaskowski07gecco,
  author = {Wojciech Jaśkowski and Krzysztof Krawiec and Bartosz Wieloch},
  title = {Genetic Programming for Cross-task Knowledge Sharing},
  booktitle = {Genetic and Evolutionary Computation Conference GECCO},
  publisher = {Association for Computing Machinery},
  year = {2007},
  pages = {1620--1627},
  url = {http://www.cs.put.poznan.pl/kkrawiec/pubs/p1620.pdf}
}
Jaśkowski, W., Krawiec, K. and Wieloch, B. Evolutionary Learning with Cross-Class Knowledge Reuse for Handwritten Character Recognition 2007 Proceedings of Planning to Learn Workshop, PlanLearn'07, pp. 21 - 30  inproceedings URL 
Abstract: We propose a learning algorithm that reuses knowledge acquired in
past learning sessions to improve its performance on a new learning
task. The method concerns visual learning and uses genetic programming
to represent hypotheses, each of them being a procedure that processes
visual primitives derived from the training images. The process of
recognition is generative, i.e., a procedure is supposed to restore
the shape of the processed object by drawing its reproduction on
a separate canvas. This basic method is extended with a knowledge
reuse mechanism that allows learners to import genetic material from
hypotheses that evolved for the other decision classes (object classes).
We compare both methods on a task of handwritten character recognition,
and conclude that knowledge reuse leads to significant improvement
of classification accuracy and reduces the risk of overfitting.
BibTeX:
@inproceedings{Jaskowski07PlanLearn,
  author = {Wojciech Jaśkowski and Krzysztof Krawiec and Bartosz Wieloch},
  title = {Evolutionary Learning with Cross-Class Knowledge Reuse for Handwritten Character Recognition},
  booktitle = {Proceedings of Planning to Learn Workshop, PlanLearn'07},
  year = {2007},
  pages = {21 -- 30},
  url = {http://www.cs.put.poznan.pl/kkrawiec/pubs/2007ECMLPlanLearn.pdf}
}
Jaśkowski, W., Krawiec, K. and Wieloch, B. BrilliANT: The Winner Entry of the GECCO'2007 Ant Wars Contest 2007 (RA-05/07)  techreport URL 
Abstract: BriliANT is the name of a program that won the GECCO'2007 Ant Wars
contest. Ant Wars was one of the competitions organized within GECCO'2007,
Genetic and Evolutionary Computation Conference, in London, England,
July 7--12, 2007. The goal was to evolve a controller for a virtual
ant that collects food in a square toroidal grid environment in the
presence of a competing ant. The report describes the method and
technology we used to evolve BriliANT.
BibTeX:
@techreport{Jaskowski07AntWars,
  author = {Wojciech Jaśkowski and Krzysztof Krawiec and Bartosz Wieloch},
  title = {BrilliANT: The Winner Entry of the GECCO'2007 Ant Wars Contest},
  year = {2007},
  number = {RA-05/07},
  url = {http://www.cs.put.poznan.pl/kkrawiec/pubs/2007GECCOAnts.pdf}
}
Jaśkowski, W., Krawiec, K. and Wieloch, B. AntWars Applet 2007   misc URL 
BibTeX:
@misc{jaskowski07applet,
  author = {Wojciech Jaśkowski and Krzysztof Krawiec and Bartosz Wieloch},
  title = {AntWars Applet},
  year = {2007},
  url = {http://www.cs.put.poznan.pl/kkrawiec/antwars/}
}
Krawiec, K. Generative Learning of Visual Concepts using Multiobjective Genetic Programming 2007 Pattern Recognition Letters
Vol. 28, pp. 2385-2400 
article URL 
Abstract: This paper introduces a novel method of visual learning based on Genetic
Programming, which evolves a population of individuals (image analysis
programs) that process attributed visual primitives derived from
raw raster images. The goal is to evolve an image analysis program
that correctly recognizes the training concept (shape). The approach
uses generative evaluation scheme: individuals are rewarded for re-
producing the shape of the object being recognized using graphical
primitives and elementary background knowledge encoded in predefined
operators. Evolutionary run is driven by a multiobjective fitness
function to prevent premature convergence and enable effective exploration
of the space of solutions. We present the method in detail and verify
it experimentally on the task of learning two visual concepts from
examples.
BibTeX:
@article{Krawiec07PRL,
  author = {Krzysztof Krawiec},
  title = {Generative Learning of Visual Concepts using Multiobjective Genetic Programming},
  journal = {Pattern Recognition Letters},
  year = {2007},
  volume = {28},
  pages = {2385-2400},
  note = {DOI: 10.1016/j.patrec.2007.08.001},
  url = {http://www.cs.put.poznan.pl/kkrawiec/pubs/2006PRL.pdf}
}
Krawiec, K. Encyclopedia of Machine Learning 2007   inbook URL 
BibTeX:
@inbook{Krawiec2007EML,
  author = {Krzysztof Krawiec},
  title = {Encyclopedia of Machine Learning},
  publisher = {Springer-Verlag},
  year = {2007},
  note = {Top 5 percent of most downloaded Springer books},
  url = {http://www.amazon.com/Encyclopedia-of-Machine-Learning/dp/0387307680}
}
Krawiec, K. and Bhanu, B. Visual Learning by Evolutionary and Coevolutionary Feature Synthesis 2007 IEEE Transactions on Evolutionary Computation
Vol. 11, pp. 635-650 
article URL 
Abstract: In this paper, we present a novel method for learning complex concepts/hypotheses
directly from raw training data. The task addressed here concerns
data-driven synthesis of recognition procedures for real-world object
recognition. The method uses linear genetic programming to encode
potential solutions expressed in terms of elementary operations,
and handles the complexity of the learning task by applying cooperative
coevolution to decompose the problem automatically at the genotype
level. The training coevolves feature extraction procedures, each
being a sequence of elementary image processing and computer vision
operations applied to input images. Extensive experimental results
show that the approach attains competitive performance for 3D object
recognition in real synthetic aperture radar (SAR) imagery.
BibTeX:
@article{Krawiec07a,
  author = {Krzysztof Krawiec and Bir Bhanu},
  title = {Visual Learning by Evolutionary and Coevolutionary Feature Synthesis},
  journal = {IEEE Transactions on Evolutionary Computation},
  year = {2007},
  volume = {11},
  pages = {635-650},
  note = {DOI: 10.1109/TEVC.2006.887351},
  url = {http://ieeexplore.ieee.org/iel5/4235/4336114/04336120.pdf}
}
Krawiec, K., Howard, D. and Zhang, M. Overview of Object Detection and Image Analysis by Means of Genetic Programming Techniques 2007 Proceedings of Frontiers in the Convergence of Bioscience and Information Technologies 2007 (FBIT2007), Jeju, Korea, October 11-13, 2007, pp. 779-784  inproceedings DOI URL 
Abstract: This paper reviews the existing work in genetic programming for object
detection and image analysis. It shortly introduces the reader into
the fundamentals of evolutionary computation and presents the basics
of the genetic programming paradigm (GP), providing a rationale for
the use of GP within computer vision and pattern recognition, particularly
when applied to object detection and image analysis. It reviews the
past research on GP for vision, referring to real-world applications
where possible. It outlines possible further research directions.
BibTeX:
@inproceedings{KrawiecHowardZhang07,
  author = {Krzysztof Krawiec and Daniel Howard and Mengjie Zhang},
  title = {Overview of Object Detection and Image Analysis by Means of Genetic Programming Techniques},
  booktitle = {Proceedings of Frontiers in the Convergence of Bioscience and Information Technologies 2007 (FBIT2007), Jeju, Korea, October 11-13, 2007},
  publisher = {IEEE CS Press},
  year = {2007},
  pages = {779--784},
  url = {http://www.cs.put.poznan.pl/kkrawiec/pubs/2007FBIT.pdf},
  doi = {http://doi.org/10.1109/FBIT.2007.148}
}
Li, R., Bhanu, B. and Krawiec, K. On the Number of Subpopulations in Coevolutionary Computation: A Database Application 2007 Genetic and Evolutionary Computation Conference GECCO, pp. 489-489  inproceedings  
Abstract: Among the existing feature selection/synthesis approaches, Coevolutionary
Feature Synthesis (CFS) based on Coevolutionary Genetic Programming
(CGP) has shown good performance on a variety of applications. In
this paper, we propose an MDL-based fitness function to help pick
a reasonable number of synthesized features which is equal to the
number of subpopulations. It naturally balances the feature transformation
complexity and classification performance. Experiments on a real
image database show that the new fitness function solves the problem
quite well.
BibTeX:
@inproceedings{LiBhanuKrawiec07geccoPoster,
  author = {Rui Li and Bir Bhanu and Krzysztof Krawiec},
  title = {On the Number of Subpopulations in Coevolutionary Computation: A Database Application},
  booktitle = {Genetic and Evolutionary Computation Conference GECCO},
  publisher = {Association for Computing Machinery},
  year = {2007},
  pages = {489--489}
}
Li, R., Bhanu, B. and Krawiec, K. Hybrid Coevolutionary Algorithms vs. SVM Algorithms 2007 Genetic and Evolutionary Computation Conference GECCO, pp. 456-463  inproceedings  
Abstract: As a learning method support vector machine is regarded as one of
the best classifiers with a strong mathematical foundation. On the
other hand, evolutionary computational technique is characterized
as a soft computing learning method with its roots in the theory
of evolution. During the past decade, SVM has been commonly used
as a classifier for various applications. The evolutionary computation
has also attracted a lot of attention in pattern recognition and
has shown significant performance improvement on a variety of applications.
However, there has been no comparison of the two methods. In this
paper, first we propose an improvement of a coevolutionary computational
classification algorithm, called Improved Coevolutionary Feature
Synthesized EM (I-CFS-EM) algorithm. It is a hybrid of coevolutionary
genetic programming and EM algorithm applied on partially labeled
data. It requires less labeled data and it makes the test in a lower
dimension, which speeds up the testing. Then, we provide a comprehensive
comparison between SVM with different kernel functions and I-CFS-EM
on several real datasets. This comparison shows that I-CFS-EM outperforms
SVM in the sense of both the classification performance and the computational
efficiency in the testing phase. We also give an intensive analysis
of the pros and cons of both approaches.
BibTeX:
@inproceedings{LiBhanuKrawiec07gecco,
  author = {Rui Li and Bir Bhanu and Krzysztof Krawiec},
  title = {Hybrid Coevolutionary Algorithms vs. SVM Algorithms},
  booktitle = {Genetic and Evolutionary Computation Conference GECCO},
  publisher = {Association for Computing Machinery},
  year = {2007},
  pages = {456--463}
}
Krawiec, K. Learning High-Level Visual Concepts Using Attributed Primitives and Genetic Programming 2006 EvoWorkshops 2006, pp. 515-519  inproceedings URL 
Abstract: In this paper, we present a novel approach to genetic learning of
high-level visual concepts that works with sets of attributed visual
primitives rather than with raster images. The paper presents the
approach in detail and verifies it in an experiment concerning locating
objects in real-world 3D scenes.
BibTeX:
@inproceedings{Krawiec06a,
  author = {Krzysztof Krawiec},
  title = {Learning High-Level Visual Concepts Using Attributed Primitives and Genetic Programming},
  booktitle = {EvoWorkshops 2006},
  publisher = {Springer-Verlag},
  year = {2006},
  pages = {515--519},
  url = {http://dx.doi.org/10.1007/11732242_48}
}
Krawiec, K. Genetic Programming for Primitive-based Acquisition of Visual Concepts 2006
Vol. 156Evolutionary Computation and Global Optimization 2006 Evolutionary Computation and Global Optimization 2006, Murzasichle, Poland, May 31 - June 2, 2006. Proceedings, pp. 211 - 222 
inproceedings URL 
Abstract: We describe a novel method for acquisition of higher-level visual
concepts using GP-based learners that process attributed visual primitives
derived from raw raster images. The approach uses an original evaluation
scheme: individuals-learners are rewarded for being able to restore
the essential features (here: shape) of the visual stimulus. The
approach is general and does not require any a priori knowledge about
the particular application or target concept to be learned; the only
prerequisite is universal knowledge related to interpretation of
visual information, encoded in nodes of GP trees. The paper demonstrates
the performance of the method on a specific visual task of acquiring
the concept of a triangle from examples given in a form of raw raster
images.
BibTeX:
@inproceedings{Krawiec06,
  author = {Krzysztof Krawiec},
  title = {Genetic Programming for Primitive-based Acquisition of Visual Concepts},
  booktitle = {Evolutionary Computation and Global Optimization 2006 Evolutionary Computation and Global Optimization 2006, Murzasichle, Poland, May 31 - June 2, 2006. Proceedings},
  publisher = {Oficyna Wydawnicza Politechniki Warszawskiej},
  year = {2006},
  volume = {156},
  pages = {211 - 222},
  url = {http://www.cs.put.poznan.pl/kkrawiec/pubs/2006KAEiOG.pdf}
}
Krawiec, K. Evolutionary Learning of Primitive-Based Visual Concepts 2006 Proc. IEEE Congress on Evolutionary Computation, Sheraton Vancouver Wall Centre Hotel, Vancouver, BC, Canada July 16-21, pp. 4451-4458  inproceedings URL 
Abstract: The paper presents a novel method of evolutionary learning dedicated
to acquisition of visual concepts. The learning process takes place
in a population of Genetic Programming-based learners that process
attributed visual primitives derived from raw raster images. The
approach uses an original evaluation scheme: evolving individuals-learners
are rewarded for being able to sketch the input visual stimulus.
Recognition proceeds here as an attempt of restoring essential features
of the input image. The approach is general by being based mostly
on universal vision knowledge; only very limited amount of a priori
knowledge about the particular application or target concept to be
learned is required. We explain the method in detail and verify it
experimentally on acquisition of simple visual concepts (triangle
and section) from examples.
BibTeX:
@inproceedings{Krawiec06b,
  author = {Krzysztof Krawiec},
  title = {Evolutionary Learning of Primitive-Based Visual Concepts},
  booktitle = {Proc. IEEE Congress on Evolutionary Computation, Sheraton Vancouver Wall Centre Hotel, Vancouver, BC, Canada July 16-21},
  year = {2006},
  pages = {4451--4458},
  url = {http://ieeexplore.ieee.org/iel5/11108/35623/01688460.pdf}
}
Krawiec, K. and Lijewski, P. Genetic Graph Programming for Object Detection 2006
Vol. 4029Artificial Intelligence and Soft Computing ICAISC 2006, 8th International Conference, Zakopane, Poland, June 25-29, 2006. Proceedings, pp. 804-813 
inproceedings URL 
Abstract: In this paper, we present a novel approach to learning from visual
information given in a form of raster images. The proposed learning
method uses genetic programming to synthesize an image processing
procedure that performs the desired vision task. The evolutionary
algorithm maintains a population of individuals, initially populated
with random solutions to the problem. Each individual encodes a directed
acyclic graph, with graph nodes corresponding to elementary image
processing operations (like image arithmetic, convolution filtering,
morphological operations, etc.), and graph edges representing the
data flow. Each graph contains a single input node to feed in the
input image and an output node that yields the final processing result.
This genetic learning process is driven by a fitness function that
rewards individuals for producing output that conforms the task-specific
objectives. This evaluation is carried out with respect to the training
set of images. Thanks to such formulation, the fitness function is
the only application-dependent component of our approach, which is
thus applicable to a wide range of vision tasks (image enhancement,
object detection, object tracking, etc.). The paper presents the
approach in detail and describes the computational experiment concerning
the task of object tracking in a real-world video sequence.
BibTeX:
@inproceedings{KrawiecLijewski2006,
  author = {Krzysztof Krawiec and Patryk Lijewski},
  title = {Genetic Graph Programming for Object Detection},
  booktitle = {Artificial Intelligence and Soft Computing ICAISC 2006, 8th International Conference, Zakopane, Poland, June 25-29, 2006. Proceedings},
  publisher = {Springer-Verlag},
  year = {2006},
  volume = {4029},
  pages = {804--813},
  url = {http://dx.doi.org/10.1007/11785231_84}
}
Krawiec, K. and Wyczałek, I. Supervised Road Detection Using Machine Learning Methodology 2006 Materiały Ogólnopolskiego Sympozjum Naukowego Opracowania cyfrowe w Fotogrametrii, Teledetekcji i GIS  inproceedings  
Abstract: In this paper we present a novel method of road detection in aerial
and satellite imaging. This structural method is based on the concept
of profile, meant as a local one-dimensional cross-section (cast)
of raster image. We acquire such profiles from the image at different
orientation angles and extract from them features well discriminating
road pixels from non-road pixels. In particular, we use feature definitions
tailored to road characteristics (mostly elongation); these include,
among others, mutual similarity of close and equally orientated profiles
(road continuity) and symmetry. To improve precision of analysis,
the method computes profiles using sub-pixel sampling.

The further part of processing relies on machine learning, in particular
on supervised learning from examples. The algorithm is given a training
sample of pixels, for which the decision class assignment (road,
non-road) is known. This information may be manually entered by a
decision maker (expert) by marking image regions representing road
fragments, or alternatively it may be retrieved from an appropriate
module of a geographical information system. Given that information,
the algorithm acquires the knowledge from training examples performing
so-called inductive learning. That knowledge may be then used to
classify the remaining image pixels, for which the decision class
assignment is not known. Moreover, the knowledge may be inspected
(and potentially corrected) by the decision maker, as it is expressed
in readable form of a decision tree.

The paper presents the algorithm in detail, describes its computer
implementation, and demonstrates its application to an aerial image
of urban area. The obtained results demonstrate good performance
of the method and indicate usefulness of machine learning approach
in analysis of aerial and satellite imagery.
BibTeX:
@inproceedings{KrawiecWyczalek06,
  author = {Krzysztof Krawiec and Ireneusz Wyczałek},
  title = {Supervised Road Detection Using Machine Learning Methodology},
  booktitle = {Materiały Ogólnopolskiego Sympozjum Naukowego Opracowania cyfrowe w Fotogrametrii, Teledetekcji i GIS},
  year = {2006},
  note = {in Polish}
}
Popielarski, P., Gerke, T., Krawiec, K., Ignaszak, Z. and Ciesiółka, J. Wstęp do ilościowej identyfikacji nieciągłości w odlewach na podstawie komputerowej analizy obrazu negatywów RT 2006   unpublished URL 
BibTeX:
@unpublished{PopielarskiGerkeKrawiecIgnaszakCiesiolka06,
  author = {Paweł Popielarski and Tomasz Gerke and Krzysztof Krawiec and Zenon Ignaszak and Joanna Ciesiółka},
  title = {Wstęp do ilościowej identyfikacji nieciągłości w odlewach na podstawie komputerowej analizy obrazu negatywów RT},
  year = {2006},
  note = {XI Sympozjum Modelowanie Procesów Odlewniczych, Ostrowieczno, 26-27.10.2006},
  url = {http://www.mat.put.poznan.pl/ostrowieczno/}
}
Bhanu, B., Lin, Y. and Krawiec, K. Evolutionary Synthesis of Pattern Recognition Systems 2005   book URL 
BibTeX:
@book{BhanuLinKrawiec05,
  author = {Bir Bhanu and Yingqiang Lin and Krzysztof Krawiec},
  title = {Evolutionary Synthesis of Pattern Recognition Systems},
  publisher = {Springer-Verlag},
  year = {2005},
  url = {http://www.springer.com/west/home/computer/imaging?SGWID=4-149-22-39144807-detailsPage=ppmmedia|aboutThisBook}
}
Krawiec, K. Koewolucja kooperatywna w dekompozycji zadań rozpoznawania obrazów 2005
Vol. 2Sztuczna inteligencja w zarządzaniu i sterowaniu, pp. 333-358 
incollection  
BibTeX:
@incollection{Krawiec05,
  author = {Krzysztof Krawiec},
  title = {Koewolucja kooperatywna w dekompozycji zadań rozpoznawania obrazów},
  booktitle = {Sztuczna inteligencja w zarządzaniu i sterowaniu},
  publisher = {Wydawnictwo Uniwersytetu Śląskiego},
  year = {2005},
  volume = {2},
  pages = {333-358},
  note = {ISSN 1732-3517}
}
Krawiec, K. and Bhanu, B. Visual Learning by Coevolutionary Feature Synthesis 2005 IEEE Transactions on System, Man, and Cybernetics -- Part B
Vol. 35(3), pp. 409-425 
article URL 
Abstract: In this paper, a novel genetically inspired visual learning method
is proposed. Given the training raster images, this general approach
induces a sophisticated feature-based recognition system. It employs
the paradigm of cooperative coevolution to handle the computational
difficulty of this task. To represent the feature extraction agents,
the linear genetic programming is used. The paper describes the learning
algorithm and provides a firm rationale for its design. Different
architectures of recognition systems are considered that employ the
proposed feature synthesis method. An extensive experimental evaluation
on the demanding real-world task of object recognition in synthetic
aperture radar (SAR) imagery shows the ability of the proposed approach
to attain high recognition performance in different operating conditions.
BibTeX:
@article{KrawiecBhanu05,
  author = {Krzysztof Krawiec and Bir Bhanu},
  title = {Visual Learning by Coevolutionary Feature Synthesis},
  journal = {IEEE Transactions on System, Man, and Cybernetics -- Part B},
  year = {2005},
  volume = {35},
  number = {3},
  pages = {409--425},
  url = {http://ieeexplore.ieee.org/iel5/3477/30862/01430827.pdf}
}
Krawiec, K. and Bhanu, B. Automatic Feature Synthesis for Object Recognition in Radar Modality – Selected Results 2005 (RA-022/2005)School: Poznan University of Technology, Institute of Computing Science  techreport  
Abstract: In this paper, we present a novel method for learning complex concepts/hypotheses directly from raw training data. The task addressed here concerns data-driven synthesis of recognition procedures for real-world object recognition. The method uses linear genetic programming to encode potential solutions expressed in terms of elementary operations, and handles the complexity of the learning task by applying cooperative coevolution to decompose the problem automatically. The training coevolves feature extraction procedures, each being a sequence of elementary image processing and feature extraction operations applied to input images. Extensive experimental results show that the approach attains competitive performance for 3‑D object recognition in real synthetic aperture radar (SAR) imagery.
BibTeX:
@techreport{Krawiec05featSynthReport,
  author = {Krzysztof Krawiec and Bir Bhanu},
  title = {Automatic Feature Synthesis for Object Recognition in Radar Modality – Selected Results},
  school = {Poznan University of Technology, Institute of Computing Science},
  year = {2005},
  number = {RA-022/2005}
}
Krawiec, K. Evolutionary Feature Programming: Cooperative learning for knowledge discovery and computer vision 2004   book URL 
Abstract: This book concerns the methodology of machine learning algorithms
that explicitly change the representation of their training data
while learning. This process, known as feature construction or transformation
of representation, rewrites learner's input data, getting rid of
useless data components, and combining the useful ones in synergetic
way with of background knowledge. The objective is to improve learner's
predictive performance and/or to enable access to input data that
is incompatible with learning algorithm, and could not be used directly
(e.g., raster images). The new methodology elaborated in this book,
termed evolutionary feature programming (EFP), puts the feature construction
task into optimization perspective and uses evolutionary computation
to effectively search the space of solutions. Each of the evolving
individuals represents a specific feature extraction procedure. We
design a novel variant of genetic programming to encode the way the
training data undergoes transformation prior to being fed into learning
algorithm. The book provides extensive rationale for this particular
design and genetic encoding of solutions. Apart from this canonical
approach, we propose a methodology for tackling with complexity of
EFP. In coevolutionary feature programming (CFP), we decompose the
feature construction task using cooperative coevolution, a variant
of evolutionary computation that allows for semi-independent elaboration
of solution components. We propose and discuss four different decomposition
strategies for breaking up the feature construction process. The
practical utility of EFP and CFP is verified in two qualitatively
different application areas: machine learning from examples given
in attribute-value form, and visual learning from raw raster images.
Considered real-world case studies concern glass type identification,
diagnosing of diabetes, sonar-based object identification, 3D object
recognition in visible spectrum, and vehicle identification in radar
modality. The experimental results indicate that the proposed methodology
is general and proves effective in different environments, and that
it exhibits features that are appealing from practical viewpoint
(performance, scalability, generalization, explanatory character,
to mention the most important ones).
BibTeX:
@book{Krawiec2004a,
  author = {Krzysztof Krawiec},
  title = {Evolutionary Feature Programming: Cooperative learning for knowledge discovery and computer vision},
  publisher = {Wydawnictwo Politechniki Poznanskiej},
  year = {2004},
  url = {http://www.cs.put.poznan.pl/kkrawiec/pubs/hab/krawiec_hab.pdf}
}
Krawiec, K., Stefanowski, J. and Słowiński, R. Multiple classifiers for learning from pictorial information 2004 7th European Congress on Telepathology, July 8-11, 2004, Poznan, Poland, pp. 46  inproceedings URL 
BibTeX:
@inproceedings{Krawiec2004,
  author = {Krzysztof Krawiec and Jerzy Stefanowski and Roman Słowiński},
  title = {Multiple classifiers for learning from pictorial information},
  booktitle = {7th European Congress on Telepathology, July 8-11, 2004, Poznan, Poland},
  year = {2004},
  pages = {46},
  note = {Abstract in Electronic Journal of Pathology and Histology},
  url = {http://ejpath.amu.edu.pl/}
}
Krawiec, K. and Włodarski, L. Coevolutionary feature construction for transformation of representation 2004 New Trends in Intelligent Information Processing and Web Mining, pp. 139-150  inproceedings  
Abstract: In this paper, a novel method of symbolic feature construction for
machine learners is proposed. The method uses evolutionary computation
to evolve feature transformation expressions, encoded in a form of
fixed-length sequences of predefined elementary operations. Two variants
of the method are proposed. One of them involves evolutionary computation
for searching the space of possible solutions, whereas the other
one engages cooperative coevolution for the same purpose. The results
of extensive experimental evaluation on reference machine learning
problems indicate superiority of the coevolutionary variety of the
proposed approach.
BibTeX:
@inproceedings{KrawiecWlodarski04,
  author = {K. Krawiec and L. Włodarski},
  title = {Coevolutionary feature construction for transformation of representation},
  booktitle = {New Trends in Intelligent Information Processing and Web Mining},
  publisher = {Springer-Verlag},
  year = {2004},
  pages = {139--150}
}
Bhanu, B. and Krawiec, K. Coevolving Feature Extraction Agents for Target Recognition in SAR Images 2003
Vol. 5095Algorithms for Synthetic Aperture Radar Imagery X, pp. 275-283 
inproceedings  
BibTeX:
@inproceedings{BhanuKrawiec03,
  author = {B. Bhanu and K. Krawiec},
  title = {Coevolving Feature Extraction Agents for Target Recognition in SAR Images},
  booktitle = {Algorithms for Synthetic Aperture Radar Imagery X},
  year = {2003},
  volume = {5095},
  pages = {275--283}
}
Krawiec, K. Computer Vision using Genetic Programming Library (CVGP) 2003 School: Center for Research in Intelligent Systems, University of California, Riverside  techreport  
Abstract: The purpose of this document is to provide a basic knowledge required to run experiments in the Computer Vision using Genetic Programming Library (CVGP) environment.
BibTeX:
@techreport{Krawiec03CVGPusersGuide,
  author = {Krzysztof Krawiec},
  title = {Computer Vision using Genetic Programming Library (CVGP)},
  school = {Center for Research in Intelligent Systems, University of California, Riverside},
  year = {2003}
}
Krawiec, K. and Bhanu, B. Visual Learning by Evolutionary Feature Synthesis 2003 Proceedings of the Twentieth International Conference on Machine Learning (ICML 2003), pp. 376-383  inproceedings  
Abstract: In this paper, we present a novel method for learning complex concepts/hypotheses
directly from raw training data. The task addressed here concerns
data-driven synthesis of recognition procedures for real-world object
recognition task. The method uses linear genetic programming to encode
potential solutions expressed in terms of elementary operations,
and handles the complexity of the learning task by applying cooperative
coevolution to decompose the problem automatically. The training
consists in coevolving feature extraction procedures, each being
a sequence of elementary image processing and feature extraction
operations. Extensive experimental results show that the approach
attains competitive performance for 3-D object recognition in real
synthetic aperture radar (SAR) imagery.
BibTeX:
@inproceedings{KrawiecBhanu03,
  author = {K. Krawiec and B. Bhanu},
  title = {Visual Learning by Evolutionary Feature Synthesis},
  booktitle = {Proceedings of the Twentieth International Conference on Machine Learning (ICML 2003)},
  publisher = {AAAI Press},
  year = {2003},
  pages = {376--383}
}
Krawiec, K. and Bhanu, B. Coevolutionary Feature Learning for Object Recognition 2003
Vol. 2734Machine Learning and Data Mining in Pattern Recognition. Third International Conference, pp. 224-238 
inproceedings  
BibTeX:
@inproceedings{KrawiecBhanu03c,
  author = {K. Krawiec and B. Bhanu},
  title = {Coevolutionary Feature Learning for Object Recognition},
  booktitle = {Machine Learning and Data Mining in Pattern Recognition. Third International Conference},
  publisher = {Springer-Verlag},
  year = {2003},
  volume = {2734},
  pages = {224--238}
}
Krawiec, K. and Bhanu, B. Coevolutionary Computation for Synthesis of Recognition Systems 2003 Proceedings of Computer Vision and Pattern Recognition Conference, Workshop on Learning in Computer Vision and Pattern Recognition CVPR  inproceedings  
BibTeX:
@inproceedings{KrawiecBhanu03b,
  author = {K. Krawiec and B. Bhanu},
  title = {Coevolutionary Computation for Synthesis of Recognition Systems},
  booktitle = {Proceedings of Computer Vision and Pattern Recognition Conference, Workshop on Learning in Computer Vision and Pattern Recognition CVPR},
  year = {2003}
}
Krawiec, K. and Bhanu, B. Coevolution and Linear Genetic Programming for Visual Learning 2003
Vol. 2723Genetic and Evolutionary Computation, pp. 332-343 
inproceedings  
Abstract: In this paper, a novel genetically-inspired visual learning method
is proposed. Given the training images, this general approach induces
a sophisticated feature-based recognition system, by using cooperative
coevolution and linear genetic programming for the procedural representation
of feature extraction agents. The paper describes the learning algorithm
and provides a firm rationale for its design. An extensive experimental
evaluation, on the demanding real-world task of object recognition
in synthetic aperture radar (SAR) imagery, shows the competitiveness
of the proposed approach with human-designed recognition systems.
BibTeX:
@inproceedings{KrawiecBhanu03a,
  author = {K. Krawiec and B. Bhanu},
  title = {Coevolution and Linear Genetic Programming for Visual Learning},
  booktitle = {Genetic and Evolutionary Computation},
  publisher = {Springer-Verlag},
  year = {2003},
  volume = {2723},
  pages = {332--343}
}
Krawiec, K. and Stefanowski, J. Sieci neuronowe i uczenie maszynowe 2003   book  
BibTeX:
@book{KrawiecStefanowski03,
  author = {K. Krawiec and J. Stefanowski},
  title = {Sieci neuronowe i uczenie maszynowe},
  publisher = {Wydawnictwo Politechniki Poznańskiej},
  year = {2003}
}
Bhanu, B. and Krawiec, K. Coevolutionary Construction of Features for Transformation of Representation in Machine Learning 2002 Proc. Bird of a Feather Workshops, Genetic and Evolutionary Computation Conference, pp. 249-254  inproceedings  
BibTeX:
@inproceedings{BhanuKrawiec02,
  author = {B. Bhanu and K. Krawiec},
  title = {Coevolutionary Construction of Features for Transformation of Representation in Machine Learning},
  booktitle = {Proc. Bird of a Feather Workshops, Genetic and Evolutionary Computation Conference},
  publisher = {AAAI Press},
  year = {2002},
  pages = {249--254}
}
Krawiec, K. Genetic Programming-based Construction of Features for Machine Learning and Knowledge Discovery Tasks 2002 Genetic Programming and Evolvable Machines
Vol. 4, pp. 329-343 
article  
Abstract: In this paper we use genetic programming for changing the representation
of the input data for machine learners. In particular, the topic
of interest is here feature construction in the learning-from-examples
paradigm, where new features are built based on the original set
of attributes. The paper first introduces the general framework for
GP-based feature construction. Then, an extended approach is proposed
where the useful components of representation (features) are preserved
during an evolutionary run, as opposed to the standard approach where
valuable features are often lost during search. Finally, we present
and discuss the results of an extensive computational experiment
carried out on several reference data sets. The outcomes show that
classifiers induced using the representation enriched by the GP-constructed
features provide better accuracy of classification on the test set.
In particular, the extended approach proposed in the paper proved
to be able to outperform the standard approach on some benchmark
problems on a statistically significant level.
BibTeX:
@article{Krawiec02,
  author = {K. Krawiec},
  title = {Genetic Programming-based Construction of Features for Machine Learning and Knowledge Discovery Tasks},
  journal = {Genetic Programming and Evolvable Machines},
  year = {2002},
  volume = {4},
  pages = {329--343}
}
Krawiec, K. Pairwise comparison of hypotheses in evolutionary learning 2001 Proc. Eighteenth International Conference on Machine Learning, pp. 266-273  inproceedings  
Abstract: This paper investigates the use of evolutionary algorithms for the
search of hypothesis space in machine learning tasks. As opposed
to the common scalar evaluation function imposing a complete order
onto the hypothesis space, we propose genetic search incorporating
pairwise comparison of hypotheses. Particularly, we allow incomparability
of hypotheses, what implies a partial order in the hypothesis space.
We claim that such an extension protects the 'interesting’ hypotheses
from being discarded in the search process, and thus increases the
diversity of the population, allowing better exploration of the solution
space. As a result it is more probable to reach hypotheses with good
predictive accuracy. This supposition has been positively verified
in an extensive comparative experiment of evolutionary visual learning
concerning the recognition of handwritten characters.
BibTeX:
@inproceedings{Krawiec01,
  author = {Krzysztof Krawiec},
  title = {Pairwise comparison of hypotheses in evolutionary learning},
  booktitle = {Proc. Eighteenth International Conference on Machine Learning},
  publisher = {Morgan Kaufmann},
  year = {2001},
  pages = {266--273}
}
Krawiec, K. Pairwise Comparison of Hypotheses in Evolutionary Learning 2001 Machine Learning. Proceedings of the Eighteenth International Conference (ICML 2001, pp. 266-273  inproceedings  
BibTeX:
@inproceedings{Krawiec01pairwisecomparison,
  author = {Krzysztof Krawiec},
  title = {Pairwise Comparison of Hypotheses in Evolutionary Learning},
  booktitle = {Machine Learning. Proceedings of the Eighteenth International Conference (ICML 2001},
  publisher = {Morgan Kaufmann Publishers},
  year = {2001},
  pages = {266--273}
}
Krawiec, K. Pairwise Comparison of Hypotheses Coverings as a Natural Mean Against Undesirable Niching in Evolutionary Inductive Learning 2001 (RA-005/2001)School: Poznan University of Technology, Institute of Computing Science  techreport  
Abstract: This report summarizes the results of research on the use of evolutionary learning for solving pattern recognition problems. The general idea consists in evolutionary search in the space of pattern recognition programs. The whole body of results described here was obtained in the improved version of GPVIS environment [15]. In particular, this report describes the 2.0 version of the environment and is devoted in a great part to the extensions beyond the standard genetic programming introduced into GPVIS, including the novel method of hypothesis evaluation proposed for evolutionary learning.
This work focuses on reasoning from pictorial information based on evolutionary computation, or, to be more precise, on the paradigm of genetic programming [12]. The outline of the method is as follows. The genetic search engine performs the search through the space of image processing and analysis programs. The programs have the form of expressions formulated in a specialized language called GPVISL (Genetic Programming for Visual Learning language). The genetic search engine realizes the selection of parent solutions (individuals), which are then crossed over and mutated to obtain the next generation of solutions. The selection is done w.r.t. the value of evaluation (fitness) function. A solution is evaluated by testing its behavior on a set of fitness cases, which are equivalent to images in this context. The fitness function is the percentage of 'hits' [14], i.e. of the correct decisions (recognitions) made by the system.
BibTeX:
@techreport{Krawiec01pairwiseReport,
  author = {Krzysztof Krawiec},
  title = {Pairwise Comparison of Hypotheses Coverings as a Natural Mean Against Undesirable Niching in Evolutionary Inductive Learning},
  school = {Poznan University of Technology, Institute of Computing Science},
  year = {2001},
  number = {RA-005/2001}
}
Krawiec, K. On the Use of Pairwise Comparison of Hypotheses in Evolutionary Learning Applied to Learning from Visual Examples 2001
Vol. 2123Machine Learning and Data Mining in Pattern Recognition, pp. 307-321 
inproceedings  
BibTeX:
@inproceedings{Krawiec01b,
  author = {K. Krawiec},
  title = {On the Use of Pairwise Comparison of Hypotheses in Evolutionary Learning Applied to Learning from Visual Examples},
  booktitle = {Machine Learning and Data Mining in Pattern Recognition},
  publisher = {Springer-Verlag},
  year = {2001},
  volume = {2123},
  pages = {307--321}
}
Krawiec, K. Genetic Programming with Local Improvement for Visual Learning from Examples 2001
Vol. 2124Computer Analysis of Images and Patterns, pp. 209-216 
inproceedings  
BibTeX:
@inproceedings{Krawiec01a,
  author = {K. Krawiec},
  title = {Genetic Programming with Local Improvement for Visual Learning from Examples},
  booktitle = {Computer Analysis of Images and Patterns},
  publisher = {Springer-Verlag},
  year = {2001},
  volume = {2124},
  pages = {209--216}
}
Krawiec, K. Genetic programming using partial order of solutions for pattern recognition tasks 2001 Proc. II National Conference 'Computer Recognition Systems', pp. 427-433  inproceedings  
BibTeX:
@inproceedings{Krawiec01d,
  author = {K. Krawiec},
  title = {Genetic programming using partial order of solutions for pattern recognition tasks},
  booktitle = {Proc. II National Conference 'Computer Recognition Systems'},
  year = {2001},
  pages = {427--433}
}
Krawiec, K. Evolutionary Computation Framework for Learning from Visual Examples 2001 Image Processing and Communications
Vol. 7(3/4/2004), pp. 85-96 
article  
Abstract: This paper investigates the use of evolutionary programming for the
search of hypothesis space in visual learning tasks. The general
goal of the project is to elaborate human-competitive procedures
for pattern discrimination by means of learning based on the training
data (set of images). In particular, the topic addressed here is
the comparison between the ‘standard’ genetic programming (as defined
by Koza [13]) and the genetic programming extended by local optimization
of solutions, so-called genetic local search. The hypothesis formulated
in the paper is that genetic local search provides better solutions
(i.e. classifiers with higher predictive accuracy) than the genetic
search without that extension. This supposition was positively verified
in an extensive comparative experiment of visual learning concerning
the recognition of handwritten charac-ters.
BibTeX:
@article{Krawiec01c,
  author = {K. Krawiec},
  title = {Evolutionary Computation Framework for Learning from Visual Examples},
  journal = {Image Processing and Communications},
  year = {2001},
  volume = {7},
  number = {3/4/2004},
  pages = {85--96}
}
Komosiński, M. and Krawiec, K. Evolutionary weighting of image features for diagnosing of CNS tumors 2000 Artificial Intelligence in Medicine
Vol. 19(1), pp. 25-38 
article  
BibTeX:
@article{KomosinskiKrawiec00,
  author = {Maciej Komosiński and Krzysztof Krawiec},
  title = {Evolutionary weighting of image features for diagnosing of CNS tumors},
  journal = {Artificial Intelligence in Medicine},
  year = {2000},
  volume = {19},
  number = {1},
  pages = {25--38}
}
Krawiec, K. Constructive induction of features in decision support based on pictorial information 2000 School: Poznan University of Technology  phdthesis URL 
BibTeX:
@phdthesis{Krawiec:2000:PhD,
  author = {Krzysztof Krawiec},
  title = {Constructive induction of features in decision support based on pictorial information},
  school = {Poznan University of Technology},
  year = {2000},
  url = {http://www.cs.put.poznan.pl/kkrawiec/pubs/phd.pdf}
}
Krawiec, K. Constructive Induction of Features in Decision Support based on Pictorial Information 2000   phdthesis  
BibTeX:
@phdthesis{Krawiec00,
  author = {K. Krawiec},
  title = {Constructive Induction of Features in Decision Support based on Pictorial Information},
  year = {2000}
}
Krawiec, K. Constructive induction in learning of image representation 2000 (RA-006/2000)School: Poznan University of Technology, Institute of Computing Science  techreport  
Abstract: This report describes the results of investigations on the use of genetic programming for learning in pattern recognition problems. The general idea consists in evolutionary search in the space of pattern recognition programs. The solutions are expressed in terms of the Genetic Programming for Visual Learning language (GPVISL), described in this work.
BibTeX:
@techreport{Krawiec00CIreport,
  author = {Krzysztof Krawiec},
  title = {Constructive induction in learning of image representation},
  school = {Poznan University of Technology, Institute of Computing Science},
  year = {2000},
  number = {RA-006/2000}
}
Faifer, M., Janikow, C. and Krawiec, K. Extracting fuzzy symbolic representation from artificial neural networks 1999 Proc. 18^th International Conference of the North American Fuzzy Information Processing Society, pp. 600-604  inproceedings  
BibTeX:
@inproceedings{FaiferJanikowKrawiec99,
  author = {Maciej Faifer and Cezary Janikow and Krzysztof Krawiec},
  title = {Extracting fuzzy symbolic representation from artificial neural networks},
  booktitle = {Proc. 18^th International Conference of the North American Fuzzy Information Processing Society},
  year = {1999},
  pages = {600--604}
}
Jelonek, J., Krawiec, K. and Stefanowski, J. Comparative study of feature subset selection techniques for machine learning tasks 1998 Proc. VII International Symposium Intelligent Information Systems, pp. 68-77  inproceedings  
BibTeX:
@inproceedings{JelonekKrawiecStefanowski98,
  author = {J. Jelonek and K. Krawiec and J. Stefanowski},
  title = {Comparative study of feature subset selection techniques for machine learning tasks},
  booktitle = {Proc. VII International Symposium Intelligent Information Systems},
  year = {1998},
  pages = {68--77}
}
Jelonek, J., Krawiec, K., Słowiński, R. and Szymaś, J. Grizzly -- An image processing and analysis system oriented towards medical images 1998 Journal of Decision Systems
Vol. 7(3-4) 
article  
BibTeX:
@article{JelonekKrawiecSlowinskiSzymas98,
  author = {J. Jelonek and K. Krawiec and R. Słowiński and J. Szymaś},
  title = {Grizzly -- An image processing and analysis system oriented towards medical images},
  journal = {Journal of Decision Systems},
  year = {1998},
  volume = {7},
  number = {3--4}
}
Krawiec, K., Słowiński, R. and Vanderpooten, D. Learning of decision rules from similarity based rough approximations 1998 Rough Sets in Knowledge Discovery, pp. 37-54  incollection  
BibTeX:
@incollection{1998KrawiecSimilarity,
  author = {Krzysztof Krawiec and Roman Słowiński and Daniel Vanderpooten},
  title = {Learning of decision rules from similarity based rough approximations},
  booktitle = {Rough Sets in Knowledge Discovery},
  publisher = {Physica Verlag},
  year = {1998},
  pages = {37-54}
}
Krawiec, K. and Słowiński, R. Learning Discriminating Descriptions from Images 1997 Proc. VI International Symposium Intelligent Information Systems, pp. 118-127  inproceedings  
BibTeX:
@inproceedings{KrawiecSlowinski97,
  author = {K. Krawiec and R. Słowiński},
  title = {Learning Discriminating Descriptions from Images},
  booktitle = {Proc. VI International Symposium Intelligent Information Systems},
  year = {1997},
  pages = {118--127}
}
Jelonek, J., Krawiec, K. and Słowiński, R. Rough Set Reduction of Attributes and their Domains for Neural Networks 1995 Computational Intelligence
Vol. 11(2), pp. 339-347 
article  
BibTeX:
@article{JelonekKrawiecSlowinski95,
  author = {J. Jelonek and K. Krawiec and R. Słowiński},
  title = {Rough Set Reduction of Attributes and their Domains for Neural Networks},
  journal = {Computational Intelligence},
  year = {1995},
  volume = {11},
  number = {2},
  pages = {339--347}
}