MemoryContainer
is used to store data in decision table for modifications.
It provides direct access to each field in the table, so remember that original values
are going to be modificated. The data stored in memory can be reach by giving
index to the example and then to the field. You are able to make conversions
to another data format using serial input and output.
Class | Description |
---|---|
AllDirectionsMemoryContainer |
Extention of
MemoryContainer that can produce memory container with all possible orders of preferences for all attributes. |
AttributesMeaningsDescriptions |
Class used to store descriptions of the meanings of attributes from memory container.
|
MemoryContainer |
This class is used to store examples and attributes directly into memory.
|
MemoryContainerAttrManager |
Class used to adapt
MemoryContainer as long as its implementation of getAttributes
and getAttrCount methods works wrong when there are no attributes added to memory container. |
MemoryContainerDecisionsManager |
Class used to manage decisions from given memory container.
|
MemoryContainerDescription |
Class used to store description of one memory container.
It stores: - memory container type (currently one of INFORMATION_TABLE , PCT or SIMILARITY_TABLE ),- evaluation space in which evaluations from memory container are defined (currently PARETO or LORENZ ),- reference to the AttributesMeaningsDescriptions object containing descriptions of meanings of all
attributes from memory container.Optionally it may also store one or both of the following references: - reference to the OrdinalValue2CardinalValue object containing information about how to convert
ordinal values from memory container to cardinal values,- reference to the EvaluationDifference2PreferenceIntensityDegree object containing information
about how to convert differences of evaluations on conditional cardinal criteria from memory container
to the degrees of the intensity of preference (used to transform information table to PCT).If both references are stored it means, that first conversion stored in OrdinalValue2CardinalValue
object should be applied to memory container and then conversion stored in EvaluationDifference2PreferenceIntensityDegree
should be applied to new memory container resulting from the first conversion.Predicted usage of information stored in this class is: MemoryContainer newInformationTable = ordinalValue2CardinalValue.applyTo(informationTable) ;MemoryContainer pct = InformationTable2PCTConverter.convert(newInformationTable, evaluationDifference2PreferenceIntensityDegree) ;Main advantage of this solution is that it is possible to save both user-defined conversions (ordinal values -> cardinal values + differences of evaluations on cardinal criteria -> degrees of the intensity of preference) along with only one information table without need of saving intermediate information table which may be not attractive to the user. |
MemoryContainersDescriptions |
Class used to store global descriptions for different memory containers.
|
MemoryContainersUniqueIdentifiers | Deprecated.
see
MemoryContainer.setId(String) and MemoryContainer.getId() methods |
MemoryInput |
Provides serial access to data in memory container.
|
MemoryOutput |
Makes memory container to behave as serial output.
|
RandomizableMemoryContainer |
Memory container that can be randomized.
|
TestMemoryContainerDeliverer |
Class used to store and deliver one test memory container (decision table or PCT), used by a classifier.
|
WekaTransferableMemoryContainer |
Extention of
MemoryContainer that can produce Weka's Instances object. |