org.openimaj.math.matrix.algorithm.whitening

Class ZCAWhitening

java.lang.Object

org.openimaj.math.matrix.algorithm.whitening.WhiteningTransform

org.openimaj.math.matrix.algorithm.whitening.PCAWhitening

org.openimaj.math.matrix.algorithm.whitening.ZCAWhitening

@Reference(type=Article,
           author={"Anthony J. Bell","Terrence J. Sejnowski"},
           title="The `Independent Components\' of Natural Scenes are Edge Filters.",
           year="1997",
           journal="VISION RESEARCH",
           pages={"3327","","3338"},
           volume="37")
public class ZCAWhitening
extends PCAWhitening

The ZCA Whitening transform. Works like PCA whitening, but after variance normalisation, rotates the data back to the original orientation. The benefit is that the data will still "look-like" the original input to some extent.

Author:

Jonathon Hare (jsh2@ecs.soton.ac.uk)

Field Summary

Fields inherited from class org.openimaj.math.matrix.algorithm.whitening.PCAWhitening

eps, ndims, ns, transform

Constructor Summary

Constructors

Constructor and Description
ZCAWhitening(double eps, Normaliser ns) Construct with the given variance regularization parameter and data normalisation strategy.

Method Summary

Modifier and Type	Method and Description
void	train(double[][] data) Train the whitening transform with the given features.

Methods inherited from class org.openimaj.math.matrix.algorithm.whitening.PCAWhitening

getTransform, whiten

Methods inherited from class org.openimaj.math.matrix.algorithm.whitening.WhiteningTransform

whiten

Methods inherited from class java.lang.Object

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

Constructor Detail

ZCAWhitening

public ZCAWhitening(double eps,
                    Normaliser ns)

Construct with the given variance regularization parameter and data normalisation strategy.

Parameters:

eps - the variance normalisation regularizer (each principle dimension is divided by sqrt(lamba + eps), where lamba is the corresponding eigenvalue).

ns - the normalisation to apply to each input data vector prior to training the transform or applying the actual whitening.

Method Detail

train

public void train(double[][] data)

Description copied from class: WhiteningTransform

Train the whitening transform with the given features.

Overrides:

train in class PCAWhitening

Parameters:

data - the data (one feature per row)