gov.sandia.cognition.learning.function.categorization

Class KernelBinaryCategorizer<InputType,EntryType extends WeightedValue<? extends InputType>>

java.lang.Object

gov.sandia.cognition.util.AbstractCloneableSerializable

gov.sandia.cognition.learning.function.categorization.AbstractBinaryCategorizer<InputType>

gov.sandia.cognition.learning.function.categorization.AbstractDiscriminantBinaryCategorizer<InputType>

gov.sandia.cognition.learning.function.categorization.KernelBinaryCategorizer<InputType,EntryType>

Type Parameters:

InputType - The input type for the categorizer.

EntryType - The type of weighted value entry in the categorizer's list of examples.

All Implemented Interfaces:

Evaluator<InputType,java.lang.Boolean>, BinaryCategorizer<InputType>, Categorizer<InputType,java.lang.Boolean>, DiscriminantBinaryCategorizer<InputType>, DiscriminantCategorizer<InputType,java.lang.Boolean,java.lang.Double>, ThresholdBinaryCategorizer<InputType>, KernelContainer<InputType>, CloneableSerializable, java.io.Serializable, java.lang.Cloneable

Direct Known Subclasses:

DefaultKernelBinaryCategorizer

public class KernelBinaryCategorizer<InputType,EntryType extends WeightedValue<? extends InputType>>
extends AbstractDiscriminantBinaryCategorizer<InputType>
implements KernelContainer<InputType>, ThresholdBinaryCategorizer<InputType>

The KernelBinaryCategorizer class implements a binary categorizer that uses a kernel to do its categorization. It is parameterized by a kernel function, a list of examples and their weights, and a bias term. This type of classifier represents what is learned by a standard Support Vector Machine or the Kernel Perceptron.

Since:

2.0

Author:

Justin Basilico

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
protected double	bias The bias term.
static double	DEFAULT_BIAS The default value for the bias is 0.0.
protected java.util.Collection<EntryType>	examples The list of weighted examples that are used for categorization.
protected Kernel<? super InputType>	kernel The internal kernel.

Fields inherited from class gov.sandia.cognition.learning.function.categorization.AbstractBinaryCategorizer

BINARY_CATEGORIES

Constructor Summary

Constructors

Constructor and Description
KernelBinaryCategorizer() Creates a new instance of KernelBinaryCategorizer.
KernelBinaryCategorizer(Kernel<? super InputType> kernel) Creates a new instance of KernelBinaryCategorizer with the given kernel.
KernelBinaryCategorizer(Kernel<? super InputType> kernel, java.util.Collection<EntryType> examples, double bias) Creates a new instance of KernelBinaryCategorizer with the given kernel, weighted examples, and bias.
KernelBinaryCategorizer(KernelBinaryCategorizer<InputType,? extends EntryType> other) Creates a new copy of a KernelBinaryCategorizer.

Method Summary

Modifier and Type	Method and Description
double	evaluateAsDouble(InputType input) Categorizes the given input vector as a double by: sum w_i * k(input, x_i)
double	getBias() Gets the bias term.
java.util.Collection<EntryType>	getExamples() Gets the list of weighted examples that categorizer is using.
Kernel<? super InputType>	getKernel() Gets the kernel.
double	getThreshold() Gets the threshold, which is the negative of the bias.
void	setBias(double bias) Sets the bias term.
void	setExamples(java.util.Collection<EntryType> examples) Sets the list of weighted examples that categorizer is using.
void	setKernel(Kernel<? super InputType> kernel) Sets the internal kernel.
void	setThreshold(double threshold) Sets the threshold, which is the negative of the bias.

Methods inherited from class gov.sandia.cognition.learning.function.categorization.AbstractDiscriminantBinaryCategorizer

evaluate, evaluateWithDiscriminant

Methods inherited from class gov.sandia.cognition.learning.function.categorization.AbstractBinaryCategorizer

getCategories

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.learning.function.categorization.DiscriminantCategorizer

evaluateWithDiscriminant

Methods inherited from interface gov.sandia.cognition.learning.function.categorization.Categorizer

getCategories

Methods inherited from interface gov.sandia.cognition.evaluator.Evaluator

evaluate

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

clone

Field Detail

DEFAULT_BIAS

public static final double DEFAULT_BIAS

The default value for the bias is 0.0.

See Also:

Constant Field Values

kernel

protected Kernel<? super InputType> kernel

The internal kernel.

examples

protected java.util.Collection<EntryType extends WeightedValue<? extends InputType>> examples

The list of weighted examples that are used for categorization.

bias

protected double bias

The bias term.

Constructor Detail

KernelBinaryCategorizer

public KernelBinaryCategorizer()

Creates a new instance of KernelBinaryCategorizer.

KernelBinaryCategorizer

public KernelBinaryCategorizer(Kernel<? super InputType> kernel)

Creates a new instance of KernelBinaryCategorizer with the given kernel.

Parameters:

kernel - The kernel to use.

KernelBinaryCategorizer

public KernelBinaryCategorizer(Kernel<? super InputType> kernel,
                               java.util.Collection<EntryType> examples,
                               double bias)

Creates a new instance of KernelBinaryCategorizer with the given kernel, weighted examples, and bias.

Parameters:

kernel - The kernel to use.

examples - The weighted examples.

bias - The bias.

KernelBinaryCategorizer

public KernelBinaryCategorizer(KernelBinaryCategorizer<InputType,? extends EntryType> other)

Creates a new copy of a KernelBinaryCategorizer.

Parameters:

other - The KernelBinaryCategorizer to copy.

Method Detail

evaluateAsDouble

public double evaluateAsDouble(InputType input)

Categorizes the given input vector as a double by: sum w_i * k(input, x_i)

Specified by:

evaluateAsDouble in interface DiscriminantBinaryCategorizer<InputType>

Parameters:

input - The input to categorize.

Returns:

The output categorization as a double where the sign is the categorization.

getThreshold

public double getThreshold()

Gets the threshold, which is the negative of the bias.

Specified by:

getThreshold in interface ThresholdBinaryCategorizer<InputType>

Returns:

The threshold, which is the negative of the bias.

setThreshold

public void setThreshold(double threshold)

Sets the threshold, which is the negative of the bias.

Specified by:

setThreshold in interface ThresholdBinaryCategorizer<InputType>

Parameters:

threshold - the threshold, which is the negative of the bias.

getExamples

public java.util.Collection<EntryType> getExamples()

Gets the list of weighted examples that categorizer is using.

Returns:

The list of weighted examples.

setExamples

public void setExamples(java.util.Collection<EntryType> examples)

Sets the list of weighted examples that categorizer is using.

Parameters:

examples - The list of weighted examples.

getBias

public double getBias()

Gets the bias term.

Returns:

bias The bias term.

setBias

public void setBias(double bias)

Sets the bias term.

Parameters:

bias - The bias term.

getKernel

public Kernel<? super InputType> getKernel()

Description copied from interface: KernelContainer

Gets the kernel.

Specified by:

getKernel in interface KernelContainer<InputType>

Returns:

Internal kernel encapsulated by the KernelContainer.

setKernel

public void setKernel(Kernel<? super InputType> kernel)

Sets the internal kernel.

Parameters:

kernel - The internal kernel.