Eksploracja i analiza danych  - Data Mining and Data Analysis

Przedmiot dla studentów Informatyki, specjalności SE (Politechnika Poznanska)

This course is taught in English language for Master Course of Software Engineering (Poznan University of Technology).

We plan to move these course materials to the e-learning SE system - so please verify the content of e-moodle cs put , in particular with respect to instructions for laboratories classes, small projects and homeworks (contat me or lab teaching assistent).

Prowadzący / Lecturer:

Class / Lab Teaching Assistant: not assigned (All lab instructions were prepared togther with dr inż. Magdalena Deckert)

Lectures are planned each Monday, CW-BT lecture room BT 126 godz, 13.30-15.00;

Lectures slides for the current academic year (based on 2019 version):

1. Introduction to KDD Process. -> this year joined with the preprocessing New lecture.
2. Pre-processing for data mining .
3. Discovery of classification trees.
4. Discovery of rules.
5. On evaluation of classifiers.
6. Naive Bayes, K-NN and other classifiers. The newest lecture Naive Bayes, K-NN version 2019; The additional lecture on Statistical Classifiers is put inside ALGODEC lecture link - see Dodatkowe wyklady.
7. See also the additional lecture on imbalanced data - this is an  updated PAN summer school lecture
8. A new lecture (2019 ed) on ensembles (bagging, boosting, random forest, stacking). You may check additional lectures on multiple classifiers is put insude ALGODEC lecture links  and a special short lecture on this topic special for you on ensembles + imbalance learning.
9. Cluster algorithms.
10. Association Rules; We added an extra lecture on rule evaluation - see additional handouts.
11. Sequence Data Mining
12. Prediction models - regression; Two lectures.
13. Basics of Artificial Neural Networks A lecture on MLP, RBF and Kohonen networks.
14. Non-parametric tests and qualitative data analyis: Selected non-parametric, ranked tests + Analysis of Questionnaires and Qualitative Data
15. Data Visualisation and Visual Data Mining
16. Summary of KDD and Buisness Advanced Data Processing Systems
 

Some other notes - Tutorials on Data Mining - ALGODEC COST project
 

Laboratory notes -- will be provided by our Lab Instructor. See the moodle system for cs put poznan.

Selected initial lab instructions ( M.Deckert's contribution to most of them is acknowledged):

• The first lab on analysing dependency of attributes and selection of attributes
1. Basics of statistical methods - Please read also statistical books
2. Laboratory instruction
3. Additional data set
• The second lab - data preprocessing
1. Reminder of methods for preprocessing - Please read also lecture slides # 2
2. Laboratory instruction pdf format
3. Churn data set - csv format
• The third lab - classification trees
1. Reminder of tree induction algorithms - Please read also lecture slides # 3 on trees and evaluation of classifiers
2. Laboratory instruction pdf format
3. Introduction to WEKA pdf format
4. UC Irvine Machine Learning Repository: look for additional files (but containing at least few hundreds of learning examples).
• The fourth lab - rules
1. Recall of rule induction algorithms - Please read also lecture slides on rules and ensembles of classifiers
2. Laboratory instruction pdf format
3. WEKA data set arff format
4. UC Irvine Machine Learning Repository: look for additional files (but containing at least few hundreds of learning examples).
• The fifth lab - case study no 1 on classification issues
1. Laboratory instruction pdf format
2. case data set arff format
3. case data set txt format
• The six lab - case study no 2 on clustering algorithms
1. Laboratory instruction and case 2 pdf format
2. Case recommendations - how to choose data set pdf format
3. Laboratory some theoretical and algorithmis comments pdf format
4. lab data set Cars txt data
5. Cars data set Statistica format
• The seven lab - association rules
1. Laboratory instruction pdf format
2. Laboratory some theoretical and algorithmis comments pdf format
3. lab data set vagetables
4. lab2 data set

Literatura - Course Books and References

• Data Mining: Concepts and Techniques, Jiawei Han, Micheline Kamber, Jian Pei, Morgan Kaufmann, 2005 (2 rozszerzone wydanie).
• Odkrywanie wiedzy z danych, Larose D. ( polskie tłumaczenie), PWN, Warszawa, 2006 = Also available in an original US version.
• Eksploracja danych, Hand D., Mannila H., Smyth P. (polskie tłumaczenie) = Also available in an original Enslish version.
• Data Mining and Analysis: Fundamental Concepts and Algorithms, Mohammed J. Zaki, Wagner Meira Jr [printed version by Cambridge Press, see also the authors' web page of the book]
• Inne pozycje wg. informacji prowadzacego / Ask the lecturer for additional books and see the first lecture slides

Ostatnia aktualizacja -- Last revision: May 28, 2019 - J.Stefanowski