public class InformationGainFunction extends SplitFunction
allSame, index, threshold
Constructor and Description |
---|
InformationGainFunction(int numOfFeatures,
Random random)
Construct an information gain split function (it needs
to be initialize with the corresponding data and indices)
|
Modifier and Type | Method and Description |
---|---|
boolean |
evaluate(Instance instance)
Evaluate a single instance based on the current
state of the split function
|
void |
init(Instances data,
ArrayList<Integer> indices)
Initialize the function given a specific data set
|
SplitFunction |
newInstance() |
public InformationGainFunction(int numOfFeatures, Random random)
numOfFeatures
- number of random features to userandom
- random number generatorpublic void init(Instances data, ArrayList<Integer> indices)
init
in class SplitFunction
data
- link to the original dataindices
- indices of the samples to usepublic boolean evaluate(Instance instance)
evaluate
in class SplitFunction
instance
- sample to evaluatepublic SplitFunction newInstance()
newInstance
in class SplitFunction
Copyright © 2015–2021 Fiji. All rights reserved.