org.tribuo.math

Class LinearParameters

java.lang.Object

org.tribuo.math.LinearParameters

All Implemented Interfaces:

Parameters

public class LinearParameters
extends Object
implements Parameters

A Parameters for producing single label linear models.

Constructor Summary

Constructors

Constructor and Description
LinearParameters(int numFeatures, int numLabels) Constructor.

Method Summary

Modifier and Type	Method and Description
Tensor[]	get() Get a reference to the underlying Tensor array.
Tensor[]	getEmptyCopy() This returns a DenseMatrix the same size as the Parameters.
DenseMatrix	getWeightMatrix() Returns the weight matrix.
Tensor[]	gradients(com.oracle.labs.mlrg.olcut.util.Pair<Double,SGDVector> score, SparseVector features) Generate the gradients for a particular feature vector given the loss and the per output gradients.
Tensor[]	merge(Tensor[][] gradients, int size) Merge together an array of gradient arrays.
SGDVector	predict(SparseVector example) Generates an unnormalised prediction by leftMultiply'ing the weights with the incoming features.
void	set(Tensor[] newWeights) Set the underlying Tensor array to newWeights.
void	update(Tensor[] gradients) Apply gradients to the parameters.

Methods inherited from class java.lang.Object

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

Constructor Detail

LinearParameters

public LinearParameters(int numFeatures,
                        int numLabels)

Constructor. The number of features and the number of outputs must be fixed and known in advance.

Parameters:

numFeatures - The number of features in the training dataset (excluding the bias).

numLabels - The number of outputs in the training dataset.

Method Detail

predict

public SGDVector predict(SparseVector example)

Generates an unnormalised prediction by leftMultiply'ing the weights with the incoming features.

Parameters:

example - A feature vector

Returns:

A DenseVector containing a score for each label.

gradients

public Tensor[] gradients(com.oracle.labs.mlrg.olcut.util.Pair<Double,SGDVector> score,
                          SparseVector features)

Generate the gradients for a particular feature vector given the loss and the per output gradients. This parameters returns a single element Tensor array.

Parameters:

score - The Pair returned by the objective.

features - The feature vector.

Returns:

A Tensor array with a single DenseSparseMatrix containing all gradients.

getEmptyCopy

public Tensor[] getEmptyCopy()

This returns a DenseMatrix the same size as the Parameters.

Specified by:

getEmptyCopy in interface Parameters

Returns:

A Tensor array containing a single DenseMatrix.

get

public Tensor[] get()

Description copied from interface: Parameters

Get a reference to the underlying Tensor array.

Specified by:

get in interface Parameters

Returns:

The parameters.

getWeightMatrix

public DenseMatrix getWeightMatrix()

Returns the weight matrix.

Returns:

The weight matrix.

set

public void set(Tensor[] newWeights)

Description copied from interface: Parameters

Set the underlying Tensor array to newWeights.

Specified by:

set in interface Parameters

Parameters:

newWeights - New parameters to store in this object.

update

public void update(Tensor[] gradients)

Description copied from interface: Parameters

Apply gradients to the parameters. Assumes that gradients is the same length as the parameters, and each Tensor is the same size as the corresponding one from the parameters.

The gradients are added to the parameters.

Specified by:

update in interface Parameters

Parameters:

gradients - A Tensor array of updates, with the length equal to Parameters.get().length.

merge

public Tensor[] merge(Tensor[][] gradients,
                      int size)

Description copied from interface: Parameters

Merge together an array of gradient arrays. Assumes the first dimension is the number of gradient arrays and the second dimension is the number of parameter Tensors.

Specified by:

merge in interface Parameters

Parameters:

gradients - An array of gradient update arrays.

size - The number of elements of gradients to merge. Allows gradients to have unused elements.

Returns:

A single Tensor array of the summed gradients.