public class SaturationFilter
extends java.lang.Object
This class implements the Gamberger's algorithm to remove class noise
Constructor and Description |
---|
SaturationFilter()
Constructor of the class
|
Modifier and Type | Method and Description |
---|---|
void |
createDatasets(java.lang.String trainIN,
java.lang.String trainOUT,
java.lang.String testIN,
java.lang.String testOUT)
It apllies the changes to remove the noise
|
java.util.Vector<java.lang.Integer> |
minimalCov(int positiveClass,
java.util.Vector<pnPair> U)
It runs the algorithm
|
void |
remove(int indexLit,
java.util.Vector<pnPair> Uaux)
It runs the algorithm
|
void |
run()
It runs the noise elimination algorithm for multiclass problems
|
boolean[] |
saturationFilter(int positiveClass)
Constructor of the class
|
double |
sumOfWeights(int indexLit,
java.util.Vector<pnPair> Uaux)
Constructor of the class
|
public void run()
It runs the noise elimination algorithm for multiclass problems
public boolean[] saturationFilter(int positiveClass)
Constructor of the class
positiveClass
- positive class id.public java.util.Vector<java.lang.Integer> minimalCov(int positiveClass, java.util.Vector<pnPair> U)
It runs the algorithm
positiveClass
- U
- public void remove(int indexLit, java.util.Vector<pnPair> Uaux)
It runs the algorithm
indexLit
- Uaux
- public double sumOfWeights(int indexLit, java.util.Vector<pnPair> Uaux)
Constructor of the class
indexLit
- Uaux
- public void createDatasets(java.lang.String trainIN, java.lang.String trainOUT, java.lang.String testIN, java.lang.String testOUT)
It apllies the changes to remove the noise
trainIN
- trainOUT
- testIN
- testOUT
-