gov.sandia.cognition.learning.algorithm.tree

Class VectorThresholdVarianceLearner

java.lang.Object

gov.sandia.cognition.util.AbstractCloneableSerializable

gov.sandia.cognition.learning.algorithm.tree.VectorThresholdVarianceLearner

All Implemented Interfaces:

BatchLearner<java.util.Collection<? extends InputOutputPair<? extends Vectorizable,java.lang.Double>>,VectorElementThresholdCategorizer>, DimensionFilterableLearner, DeciderLearner<Vectorizable,java.lang.Double,java.lang.Boolean,VectorElementThresholdCategorizer>, VectorThresholdLearner<java.lang.Double>, CloneableSerializable, java.io.Serializable, java.lang.Cloneable

public class VectorThresholdVarianceLearner
extends AbstractCloneableSerializable
implements VectorThresholdLearner<java.lang.Double>

The VectorThresholdVarianceLearner computes the best threshold over a dataset of vectors using the reduction in variance to determine the optimal index and threshold. This is an implementation of what is used in the CART regression tree algorithm.

Since:

2.0

Author:

Justin Basilico

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
static int	DEFAULT_MIN_SPLIT_SIZE The default value for the minimum split size is 1.
protected int[]	dimensionsToConsider The array of 0-based dimensions to consider in the input.
protected int	minSplitSize The threshold for allowing a split to be made, determined by how many instances fall in each left or right sides of the split.

Constructor Summary

Constructors

Constructor and Description
VectorThresholdVarianceLearner() Creates a new VectorThresholdVarianceLearner.
VectorThresholdVarianceLearner(int minSplitSize) Creates a new VectorThresholdVarianceLearner
VectorThresholdVarianceLearner(int minSplitSize, int... dimensionsToConsider) Creates a new VectorThresholdVarianceLearner.

Method Summary

Modifier and Type	Method and Description
DefaultPair<java.lang.Double,java.lang.Double>	computeBestGainThreshold(java.util.Collection<? extends InputOutputPair<? extends Vectorizable,java.lang.Double>> data, int dimension, double baseVariance) Computes the best information gain-threshold pair for the given dimension on the given data.
int[]	getDimensionsToConsider() Gets the dimensions that the learner is to consider.
int	getMinSplitSize() Gets the minimum split size.
VectorElementThresholdCategorizer	learn(java.util.Collection<? extends InputOutputPair<? extends Vectorizable,java.lang.Double>> data) Learns a VectorElementThresholdCategorizer from the given data by picking the vector element and threshold that best maximizes information gain.
void	setDimensionsToConsider(int... dimensionsToConsider) Gets the dimensions that the learner is to consider.
void	setMinSplitSize(int minSplitSize) Sets the minimum split size.

Methods inherited from class gov.sandia.cognition.util.AbstractCloneableSerializable

clone

Methods inherited from class java.lang.Object

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

Methods inherited from interface gov.sandia.cognition.util.CloneableSerializable

clone

Field Detail

DEFAULT_MIN_SPLIT_SIZE

public static final int DEFAULT_MIN_SPLIT_SIZE

The default value for the minimum split size is 1.

See Also:

Constant Field Values

minSplitSize

protected int minSplitSize

The threshold for allowing a split to be made, determined by how many instances fall in each left or right sides of the split. Both sides must have at least this number of instances. Must be positive.

dimensionsToConsider

protected int[] dimensionsToConsider

The array of 0-based dimensions to consider in the input. Null means all dimensions are considered.

Constructor Detail

VectorThresholdVarianceLearner

public VectorThresholdVarianceLearner()

Creates a new VectorThresholdVarianceLearner.

VectorThresholdVarianceLearner

public VectorThresholdVarianceLearner(int minSplitSize)

Creates a new VectorThresholdVarianceLearner

Parameters:

minSplitSize - The minimum split size. Must be positive.

VectorThresholdVarianceLearner

public VectorThresholdVarianceLearner(int minSplitSize,
                                      int... dimensionsToConsider)

Creates a new VectorThresholdVarianceLearner.

Parameters:

minSplitSize - The minimum split size. Must be positive.

dimensionsToConsider - The array of vector dimensions to consider. Null means all of them are considered.

Method Detail

learn

public VectorElementThresholdCategorizer learn(java.util.Collection<? extends InputOutputPair<? extends Vectorizable,java.lang.Double>> data)

Learns a VectorElementThresholdCategorizer from the given data by picking the vector element and threshold that best maximizes information gain.

Specified by:

learn in interface BatchLearner<java.util.Collection<? extends InputOutputPair<? extends Vectorizable,java.lang.Double>>,VectorElementThresholdCategorizer>

Parameters:

data - The data to learn from.

Returns:

The learned threshold categorizer, or none if there is no good categorizer.

computeBestGainThreshold

public DefaultPair<java.lang.Double,java.lang.Double> computeBestGainThreshold(java.util.Collection<? extends InputOutputPair<? extends Vectorizable,java.lang.Double>> data,
                                                                               int dimension,
                                                                               double baseVariance)

Computes the best information gain-threshold pair for the given dimension on the given data. It does this by sorting the data according to the dimension and then walking the sorted values to find the one that has the best threshold.

Parameters:

data - The data to use.

dimension - The dimension to compute the best threshold over.

baseVariance - The variance of the data.

Returns:

The pair containing the best information gain found along this dimension and the corresponding threshold.

getDimensionsToConsider

public int[] getDimensionsToConsider()

Description copied from interface: DimensionFilterableLearner

Gets the dimensions that the learner is to consider. Null means that all of them are included.

Specified by:

getDimensionsToConsider in interface DimensionFilterableLearner

Returns:

The array of vector dimensions to consider. Null means all of them are considered.

setDimensionsToConsider

public void setDimensionsToConsider(int... dimensionsToConsider)

Description copied from interface: DimensionFilterableLearner

Gets the dimensions that the learner is to consider. Null means that all of them are included.

Specified by:

setDimensionsToConsider in interface DimensionFilterableLearner

Parameters:

dimensionsToConsider - The array of vector dimensions to consider. Null means all of them are considered.

getMinSplitSize

public int getMinSplitSize()

Gets the minimum split size. This is the minimum number of examples that can fall on either side of the split for it to be valid. If there is not at least twice this number of examples in the input data, then no split is returned.

Returns:

The minimum split size. Must be positive.

setMinSplitSize

public void setMinSplitSize(int minSplitSize)

Sets the minimum split size. This is the minimum number of examples that can fall on either side of the split for it to be valid. If there is not at least twice this number of examples in the input data, then no split is returned.

Parameters:

minSplitSize - The minimum split size. Must be positive.