de.jungblut.classification.tree

Class DecisionTree

java.lang.Object

de.jungblut.classification.AbstractPredictor

de.jungblut.classification.AbstractClassifier

de.jungblut.classification.tree.DecisionTree

All Implemented Interfaces:

Classifier, Predictor

public final class DecisionTree
extends AbstractClassifier

A decision tree that can be used for classification with numerical or categorical features. The tree is built by maximizing information gain using the ID3 algorithm. If no featureTypes were supplied, the default is assumed to be nominal features at all feature dimensions.
Instances can be created by the static factory methods #create().

Author:

thomasjungblut

Method Summary

Modifier and Type	Method and Description
void	compileTree() Compiles this current tree representation into byte code and loads it into this class.
static DecisionTree	create()
static DecisionTree	create(FeatureType[] featureTypes) Creates a new decision tree with the given feature types.
static DecisionTree	createCompiledTree()
static DecisionTree	createCompiledTree(FeatureType[] featureTypes) Creates a new compiled decision tree with the given feature types.
static DecisionTree	deserialize(DataInput in) Reads a new tree from the given stream.
de.jungblut.math.DoubleVector	predict(de.jungblut.math.DoubleVector features) Classifies the given features.
static void	serialize(DecisionTree tree, DataOutput out) Writes the given tree to the output stream.
DecisionTree	setCompiled(boolean compiled) If set to true, this tree will be compiled after training time automatically.
DecisionTree	setFeatureTypes(FeatureType[] featureTypes) Sets the type of feature per index.
DecisionTree	setMaxHeight(int max) Sets the maximum height of this tree.
DecisionTree	setNumRandomFeaturesToChoose(int numRandomFeaturesToChoose) Sets the number of random features to choose from all features.Zero, negative numbers or numbers greater than the really available features indicate all features to be used.
DecisionTree	setSeed(long seed) Sets the seed for a random number generator if used.
void	train(de.jungblut.math.DoubleVector[] features, de.jungblut.math.DoubleVector[] outcome) Trains this classifier with the given features and the outcome.

Methods inherited from class de.jungblut.classification.AbstractClassifier

train

Methods inherited from class de.jungblut.classification.AbstractPredictor

extractPredictedClass, extractPredictedClass, predictedClass, predictedClass, predictProbability

Methods inherited from class java.lang.Object

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

Methods inherited from interface de.jungblut.classification.Predictor

asClassifier, extractPredictedClass, extractPredictedClass, predictedClass, predictedClass, predictProbability

Method Detail

train

public void train(de.jungblut.math.DoubleVector[] features,
                  de.jungblut.math.DoubleVector[] outcome)

Description copied from interface: Classifier

Trains this classifier with the given features and the outcome.

Specified by:

train in interface Classifier

Overrides:

train in class AbstractClassifier

outcome - the outcome must have classes labeled as doubles. E.G. in the binary case you have a single element and decide between 0d and 1d. In higher dimensional cases you have each of these single elements mapped to a dimension.

predict

public de.jungblut.math.DoubleVector predict(de.jungblut.math.DoubleVector features)

Description copied from interface: Predictor

Classifies the given features.

Returns:

the vector that contains an indicator at the index of the class. Usually zero or 1, in some cases it is a probability or activation value.

compileTree

public void compileTree()
                 throws Exception

Compiles this current tree representation into byte code and loads it into this class. This is considered faster, as the interpreted code can be optimized by the hotspot JVM.

Throws:

Exception - some error might happen during compilation or loading.

setFeatureTypes

public DecisionTree setFeatureTypes(FeatureType[] featureTypes)

Sets the type of feature per index. This should match the inputted number of features in the training method. If this isn't set at all, all attributes are assumed to be nominal.

Returns:

this decision tree instance.

setNumRandomFeaturesToChoose

public DecisionTree setNumRandomFeaturesToChoose(int numRandomFeaturesToChoose)

Sets the number of random features to choose from all features.Zero, negative numbers or numbers greater than the really available features indicate all features to be used.

Returns:

this decision tree instance.

setCompiled

public DecisionTree setCompiled(boolean compiled)

If set to true, this tree will be compiled after training time automatically.

Returns:

this decision tree instance.

setMaxHeight

public DecisionTree setMaxHeight(int max)

Sets the maximum height of this tree.

Returns:

this instance.

setSeed

public DecisionTree setSeed(long seed)

Sets the seed for a random number generator if used.

serialize

public static void serialize(DecisionTree tree,
                             DataOutput out)
                      throws IOException

Writes the given tree to the output stream. Note that the stream isn't closed here.

Throws:

IOException

deserialize

public static DecisionTree deserialize(DataInput in)
                                throws IOException

Reads a new tree from the given stream. Note that the stream isn't closed here.

Throws:

IOException

create

public static DecisionTree create()

Returns:

a default decision tree with all features beeing nominal.

create

public static DecisionTree create(FeatureType[] featureTypes)

Creates a new decision tree with the given feature types.

Parameters:

featureTypes - the types of the feature that must match the number of features in length.

Returns:

a default decision tree with all features beeing set to what has been configured in the given array.

createCompiledTree

public static DecisionTree createCompiledTree()

Returns:

a default compiled decision tree with all features beeing nominal.

createCompiledTree

public static DecisionTree createCompiledTree(FeatureType[] featureTypes)

Creates a new compiled decision tree with the given feature types.

Parameters:

featureTypes - the types of the feature that must match the number of features in length.

Returns:

a default decision tree with all features beeing set to what has been configured in the given array.