org.deeplearning4j.nn.learning

Class AdaGrad

java.lang.Object

org.deeplearning4j.nn.learning.AdaGrad

All Implemented Interfaces:

Serializable

public class AdaGrad
extends Object
implements Serializable

Vectorized Learning Rate used per Connection Weight Adapted from: http://xcorr.net/2014/01/23/adagrad-eliminating-learning-rates-in-stochastic-gradient-descent/

Author:

Adam Gibson

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
org.jblas.DoubleMatrix	adjustedGradient
int	cols
double	fudgeFactor
org.jblas.DoubleMatrix	gradient
org.jblas.DoubleMatrix	historicalGradient
int	rows

Constructor Summary

Constructors

Constructor and Description
AdaGrad(int rows, int cols)
AdaGrad(int rows, int cols, double gamma)

Method Summary

Methods

Modifier and Type	Method and Description
org.jblas.DoubleMatrix	getLearningRates(org.jblas.DoubleMatrix gradient) Gets feature specific learning rates Adagrad keeps a history of gradients being passed in.
double	getMasterStepSize()
boolean	isDecayLr()
void	setDecayLr(boolean decayLr)
void	setMasterStepSize(double masterStepSize)

Methods inherited from class java.lang.Object

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

Field Detail

historicalGradient

public org.jblas.DoubleMatrix historicalGradient

adjustedGradient

public org.jblas.DoubleMatrix adjustedGradient

fudgeFactor

public double fudgeFactor

gradient

public org.jblas.DoubleMatrix gradient

rows

public int rows

cols

public int cols

Constructor Detail

AdaGrad

public AdaGrad(int rows,
       int cols,
       double gamma)

AdaGrad

public AdaGrad(int rows,
       int cols)

Method Detail

getLearningRates

public org.jblas.DoubleMatrix getLearningRates(org.jblas.DoubleMatrix gradient)

Gets feature specific learning rates Adagrad keeps a history of gradients being passed in. Note that each gradient passed in becomes adapted over time, hence the name adagrad

Parameters:

gradient - the gradient to get learning rates for

Returns:

the feature specific learning rates

getMasterStepSize

public double getMasterStepSize()

setMasterStepSize

public void setMasterStepSize(double masterStepSize)

isDecayLr

public boolean isDecayLr()

setDecayLr

public void setDecayLr(boolean decayLr)