Package org.nd4j.linalg.learning.legacy

Class AdaGrad

java.lang.Object

org.nd4j.linalg.learning.legacy.AdaGrad

All Implemented Interfaces:

Serializable

public class AdaGrad
extends Object
implements Serializable

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field	Description
static double	DEFAULT_ADAGRAD_EPSILON
INDArray	historicalGradient
protected double	learningRate
protected int	numIterations
long[]	shape

Constructor Summary

Constructors

Constructor	Description
AdaGrad(double learningRate)
AdaGrad(double learningRate, double epsilon)
AdaGrad(int rows, int cols)
AdaGrad(int rows, int cols, double learningRate)
AdaGrad(long[] shape, double learningRate)

Method Summary

Modifier and Type	Method	Description
AdaGrad	createSubset(int index)
double	getGradient(double gradient, int column, long[] shape)
INDArray	getGradient(INDArray gradient, int iteration)	Gets feature specific learning rates Adagrad keeps a history of gradients being passed in.
INDArray	getGradient(INDArray gradient, int slice, long[] shape)
void	setStateViewArray(INDArray viewArray, int[] gradientShape, char gradientOrder, boolean initialize)
void	setStateViewArray(INDArray viewArray, long[] gradientShape, char gradientOrder, boolean initialize)
int	stateSizeForInputSize(int inputSize)
void	update(Object... args)

Methods inherited from class java.lang.Object

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

Field Detail

DEFAULT_ADAGRAD_EPSILON

public static final double DEFAULT_ADAGRAD_EPSILON

See Also:

Constant Field Values

historicalGradient

public INDArray historicalGradient

shape

public long[] shape

learningRate

protected double learningRate

numIterations

protected int numIterations

Constructor Detail

AdaGrad

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

Parameters:

rows -

cols -

learningRate -

AdaGrad

public AdaGrad(int rows,
               int cols)

AdaGrad

public AdaGrad(long[] shape,
               double learningRate)

AdaGrad

public AdaGrad(double learningRate)

AdaGrad

public AdaGrad(double learningRate,
               double epsilon)

Method Detail

stateSizeForInputSize

public int stateSizeForInputSize(int inputSize)

setStateViewArray

public void setStateViewArray(INDArray viewArray,
                              int[] gradientShape,
                              char gradientOrder,
                              boolean initialize)

setStateViewArray

public void setStateViewArray(INDArray viewArray,
                              long[] gradientShape,
                              char gradientOrder,
                              boolean initialize)

update

public void update(Object... args)

getGradient

public INDArray getGradient(INDArray gradient,
                            int iteration)

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 opName adagrad

Parameters:

gradient - the gradient to get learning rates for

iteration -

Returns:

the feature specific learning rates

getGradient

public double getGradient(double gradient,
                          int column,
                          long[] shape)

getGradient

public INDArray getGradient(INDArray gradient,
                            int slice,
                            long[] shape)

createSubset

public AdaGrad createSubset(int index)