public static final class DecisionTreeInfoGain.GainCriteria extends Enum
Modifier and Type | Field and Description |
---|---|
static DecisionTreeInfoGain.GainCriteria |
DEVIANCE
A measure of the quality of fit.
|
static DecisionTreeInfoGain.GainCriteria |
GINI_INDEX
A measure of statistical dispersion.
|
static DecisionTreeInfoGain.GainCriteria |
SHANNON_ENTROPY
A measure of randomness or uncertainty.
|
Modifier and Type | Method and Description |
---|---|
static DecisionTreeInfoGain.GainCriteria |
valueOf(String name) |
static DecisionTreeInfoGain.GainCriteria[] |
values() |
public static final DecisionTreeInfoGain.GainCriteria DEVIANCE
For a categorical variable having distinct values over a data set , the Deviance measure is
where and is the number of cases with on the node.public static final DecisionTreeInfoGain.GainCriteria GINI_INDEX
For a categorical variable having distinct values over a data set , the Gini index is defined as
where denotes the probability that the variable is equal to the state on the data set, .public static final DecisionTreeInfoGain.GainCriteria SHANNON_ENTROPY
For a categorical variable having distinct values over a data set , the Shannon Entropy is defined as
where and where if .public static DecisionTreeInfoGain.GainCriteria valueOf(String name)
public static DecisionTreeInfoGain.GainCriteria[] values()
Copyright © 1970-2016 Rogue Wave Software
Built May 19 2016.