de.jungblut.classification.nn

Class RBM.RBMBuilder

java.lang.Object

de.jungblut.classification.nn.RBM.RBMBuilder

Enclosing class:

RBM

public static class RBM.RBMBuilder
extends Object

Method Summary

Modifier and Type	Method and Description
RBM.RBMBuilder	batchParallelism(int numThreads)
RBM	build()
static RBM.RBMBuilder	create(ActivationFunction activation, int... layer) Creates a new RBM.RBMBuilder from an activation function and layersizes.
RBM.RBMBuilder	lambda(double lambda) Sets the regularization parameter lambda, defaults to 0 if not set.
RBM.RBMBuilder	miniBatchSize(int size)
RBM.RBMBuilder	stochastic() Sets the training mode to stochastic.
RBM.RBMBuilder	stochastic(boolean stochastic) If verbose is true, stochastic training will be used.
RBM.RBMBuilder	trainingType(TrainingType type) Sets the training type, it defaults to CPU- so only use if you want to use the GPU.
RBM.RBMBuilder	verbose() Sets verbose to true.
RBM.RBMBuilder	verbose(boolean verbose) If verbose is true, progress indicators will be printed to STDOUT.

Methods inherited from class java.lang.Object

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

Method Detail

trainingType

public RBM.RBMBuilder trainingType(TrainingType type)

Sets the training type, it defaults to CPU- so only use if you want to use the GPU.

lambda

public RBM.RBMBuilder lambda(double lambda)

Sets the regularization parameter lambda, defaults to 0 if not set.

miniBatchSize

public RBM.RBMBuilder miniBatchSize(int size)

Parameters:

size - the minibatch size to use. Batches are calculated in parallel on every cpu core if not overridden by batchParallelism(int).

batchParallelism

public RBM.RBMBuilder batchParallelism(int numThreads)

Parameters:

numThreads - set the number of threads where batches should be calculated in parallel.

verbose

public RBM.RBMBuilder verbose()

Sets verbose to true. Progress indicators will be printed to STDOUT.

stochastic

public RBM.RBMBuilder stochastic()

Sets the training mode to stochastic.

stochastic

public RBM.RBMBuilder stochastic(boolean stochastic)

If verbose is true, stochastic training will be used.

verbose

public RBM.RBMBuilder verbose(boolean verbose)

If verbose is true, progress indicators will be printed to STDOUT.

create

public static RBM.RBMBuilder create(ActivationFunction activation,
                                    int... layer)

Creates a new RBM.RBMBuilder from an activation function and layersizes.

Parameters:

activation - the activation function.

layer - an array of hidden layer sizes.

Returns:

a new RBM builder.

build

public RBM build()

Returns:

a new RBM with the given configuration.