Package unifeat.featureSelection.filter

Class FilterApproach

java.lang.Object

unifeat.featureSelection.FeatureSelection

unifeat.featureSelection.filter.FilterApproach

Direct Known Subclasses:

IRRFSACO_1, IRRFSACO_2, MGSACO, MRMR, MutualCorrelation, RRFS, RRFS, RRFSACO_1, RRFSACO_2, RSM, UFSACO

public abstract class FilterApproach
extends FeatureSelection

The abstract class contains the main methods and fields that are used in all filter-based feature selection methods.

Author:

Sina Tabakhi

See Also:

• FeatureSelection

Field Summary

Fields inherited from class unifeat.featureSelection.FeatureSelection

numClass, numFeatures, numSelectedFeature, selectedFeatureSubset, trainSet

Constructor Summary

Constructors

Constructor	Description
FilterApproach(int sizeSelectedFeatureSubset)	Initializes the parameters

Method Summary

Modifier and Type	Method	Description
static FilterApproach	newMethod(FilterType type, boolean isSupervised, Object... arguments)	This method creates new object from one of the classes that has been inherited from the FilterApproach class according to type of the feature selection method.

Methods inherited from class unifeat.featureSelection.FeatureSelection

evaluateFeatures, getSelectedFeatureSubset, loadDataSet, loadDataSet, setNumSelectedFeature, validate

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

FilterApproach

public FilterApproach(int sizeSelectedFeatureSubset)

Initializes the parameters

Parameters:

sizeSelectedFeatureSubset - the size of selected features subset

Method Details

newMethod

public static FilterApproach newMethod(FilterType type,
 boolean isSupervised,
 Object... arguments)

This method creates new object from one of the classes that has been inherited from the FilterApproach class according to type of the feature selection method.

Parameters:

type - type of the filter feature selection method

isSupervised - shows whether a method is supervised or unsupervised

arguments - a list of arguments that is applied in the feature selection method

Returns:

the created object that has been inherited from the FilterApproach class