jsat.math.optimization.stochastic

Class Rprop

java.lang.Object

jsat.math.optimization.stochastic.Rprop

All Implemented Interfaces:

Serializable, GradientUpdater

public class Rprop
extends Object
implements GradientUpdater

The Rprop algorithm provides adaptive learning rates using only first order information. Rprop works best with the true gradient, and may not work well when using stochastic gradients.

See: Riedmiller, M., & Braun, H. (1993). A direct adaptive method for faster backpropagation learning: the RPROP algorithm. IEEE International Conference on Neural Networks, 1(3), 586–591. doi:10.1109/ICNN.1993.298623

Author:

Edward Raff

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
Rprop() Creates a new Rprop instance for gradient updating
Rprop(Rprop toCopy) Copy constructor

Method Summary

Modifier and Type	Method and Description
Rprop	clone()
void	setup(int d) Sets up this updater to update a weight vector of dimension d by a gradient of the same dimension
void	update(Vec w, Vec grad, double eta) Updates the weight vector x such that x = x-ηf(grad), where f(grad) is some function on the gradient that effectively returns a new vector.
double	update(Vec w, Vec grad, double eta, double bias, double biasGrad) Updates the weight vector x such that x = x-ηf(grad), where f(grad) is some function on the gradient that effectively returns a new vector.

Methods inherited from class java.lang.Object

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

Constructor Detail

Rprop

public Rprop()

Creates a new Rprop instance for gradient updating

Rprop

public Rprop(Rprop toCopy)

Copy constructor

Parameters:

toCopy - the object to copy

Method Detail

update

public void update(Vec w,
                   Vec grad,
                   double eta)

Description copied from interface: GradientUpdater

Updates the weight vector x such that x = x-ηf(grad), where f(grad) is some function on the gradient that effectively returns a new vector. It is not necessary for the internal implementation to ever explicitly form any of these objects, so long as x is mutated to have the correct result.

Specified by:

update in interface GradientUpdater

Parameters:

w - the vector to mutate such that is has been updated by the gradient

grad - the gradient to update the weight vector x from

eta - the learning rate to apply

update

public double update(Vec w,
                     Vec grad,
                     double eta,
                     double bias,
                     double biasGrad)

Description copied from interface: GradientUpdater

Updates the weight vector x such that x = x-ηf(grad), where f(grad) is some function on the gradient that effectively returns a new vector. It is not necessary for the internal implementation to ever explicitly form any of these objects, so long as x is mutated to have the correct result.

This version of the update method includes two extra parameters to make it easer to use when a scalar bias term is also used

Specified by:

update in interface GradientUpdater

Parameters:

w - the vector to mutate such that is has been updated by the gradient

grad - the gradient to update the weight vector x from

eta - the learning rate to apply

bias - the bias term of the vector

biasGrad - the gradient for the bias term

Returns:

the value to change the bias by, the update being bias = bias - returnValue

setup

public void setup(int d)

Description copied from interface: GradientUpdater

Sets up this updater to update a weight vector of dimension d by a gradient of the same dimension

Specified by:

setup in interface GradientUpdater

Parameters:

d - the dimension of the weight vector that will be updated

clone

public Rprop clone()

Specified by:

clone in interface GradientUpdater

Overrides:

clone in class Object