jsat.classifiers.neuralnetwork.initializers

Class GaussianNormalInit

java.lang.Object

jsat.classifiers.neuralnetwork.initializers.GaussianNormalInit

All Implemented Interfaces:

java.io.Serializable, BiastInitializer, WeightInitializer

public class GaussianNormalInit
extends java.lang.Object
implements WeightInitializer, BiastInitializer

This object initializes the values of weights by sampling from the zero mean Gaussian

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
GaussianNormalInit(double stndDev) Creates a new GuassianNormalInit object for initializing weights

Method Summary

Modifier and Type	Method and Description
GaussianNormalInit	clone()
double	getStndDev()
void	init(Matrix w, java.util.Random rand) Initializes the values of the given weight matrix
void	init(Vec b, int fanIn, java.util.Random rand) Performs the initialization of the given vector of bias values
void	setStndDev(double stndDev) Sets the standard deviation of the distribution that will be sampled from

Methods inherited from class java.lang.Object

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

Constructor Detail

GaussianNormalInit

public GaussianNormalInit(double stndDev)

Creates a new GuassianNormalInit object for initializing weights

Parameters:

stndDev - the standard deviation of the distribution to sample from

Method Detail

setStndDev

public void setStndDev(double stndDev)

Sets the standard deviation of the distribution that will be sampled from

Parameters:

stndDev - the standard deviation to use

getStndDev

public double getStndDev()

Returns:

the standard deviation of the Gaussian that is sampled from

init

public void init(Matrix w,
                 java.util.Random rand)

Description copied from interface: WeightInitializer

Initializes the values of the given weight matrix

Specified by:

init in interface WeightInitializer

Parameters:

w - the matrix to initialize

rand - the source of randomness for the initialization

init

public void init(Vec b,
                 int fanIn,
                 java.util.Random rand)

Description copied from interface: BiastInitializer

Performs the initialization of the given vector of bias values

Specified by:

init in interface BiastInitializer

Parameters:

b - the vector to store the biases in

fanIn - the number of connections coming into the layer that these biases are for.

rand - the source of randomness for initialization

clone

public GaussianNormalInit clone()

Specified by:

clone in interface BiastInitializer

Specified by:

clone in interface WeightInitializer

Overrides:

clone in class java.lang.Object