org.encog.engine.network.train.prop
Class TrainFlatNetworkProp

java.lang.Object
  org.encog.engine.network.train.prop.TrainFlatNetworkProp

All Implemented Interfaces:

TrainFlatNetwork

Direct Known Subclasses:

TrainFlatNetworkBackPropagation, TrainFlatNetworkManhattan, TrainFlatNetworkResilient, TrainFlatNetworkSCG

public abstract class TrainFlatNetworkProp
extends Object
implements TrainFlatNetwork

Train a flat network using multithreading, and GPU support. The training data must be indexable, it will be broken into groups for each thread to process. At the end of each iteration the training from each thread is aggregated back to the neural network.

Field Summary

protected double	currentError The current error is the average error over all of the threads.
protected double[]	gradients The gradients.
protected FlatNetwork	network The network to train.

Constructor Summary

TrainFlatNetworkProp(FlatNetwork network, EngineDataSet training)
Train a flat network multithreaded.

Method Summary

void	calculateGradients() Calculate the gradients.
void	finishTraining() Training is to stop, free any resources.
double	getError()
int	getIteration()
double[]	getLastGradient()
FlatNetwork	getNetwork()
int	getNumThreads()
EngineDataSet	getTraining()
void	iteration() Perform one training iteration.
void	iteration(int count) Perform the specified number of training iterations.
protected void	learn() Apply and learn.
protected void	learnLimited() Apply and learn.
void	report(double[] gradients, double error, Throwable ex) Called by the worker threads to report the progress at each step.
void	setIteration(int iteration) Set the iteration.
void	setNumThreads(int numThreads) Set the number of threads to use.
abstract double	updateWeight(double[] gradients, double[] lastGradient, int index) Update a weight, the means by which weights are updated vary depending on the training.

Methods inherited from class java.lang.Object

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

Field Detail

gradients

protected double[] gradients

The gradients.

network

protected final FlatNetwork network

The network to train.

currentError

protected double currentError

The current error is the average error over all of the threads.

Constructor Detail

TrainFlatNetworkProp

public TrainFlatNetworkProp(FlatNetwork network,
                            EngineDataSet training)

Train a flat network multithreaded.

Parameters:

network - The network to train.

training - The training data to use.

Method Detail

calculateGradients

public void calculateGradients()

Calculate the gradients.

finishTraining

public void finishTraining()

Training is to stop, free any resources.

Specified by:

finishTraining in interface TrainFlatNetwork

getError

public double getError()

Specified by:

getError in interface TrainFlatNetwork

Returns:

The error from the neural network.

getLastGradient

public double[] getLastGradient()

Returns:

The gradients from the last iteration;

getNetwork

public FlatNetwork getNetwork()

Specified by:

getNetwork in interface TrainFlatNetwork

Returns:

The trained neural network.

getNumThreads

public int getNumThreads()

Specified by:

getNumThreads in interface TrainFlatNetwork

Returns:

The number of threads.

getTraining

public EngineDataSet getTraining()

Specified by:

getTraining in interface TrainFlatNetwork

Returns:

The data we are training with.

iteration

public void iteration()

Perform one training iteration.

Specified by:

iteration in interface TrainFlatNetwork

learn

protected void learn()

Apply and learn.

learnLimited

protected void learnLimited()

Apply and learn. This is the same as learn, but it checks to see if any of the weights are below the limit threshold. In this case, these weights are zeroed out. Having two methods allows the regular learn method, which is what is usually use, to be as fast as possible.

report

public void report(double[] gradients,
                   double error,
                   Throwable ex)

Called by the worker threads to report the progress at each step.

Parameters:

gradients - The gradients from that worker.

error - The error for that worker.

ex - The exception.

setNumThreads

public void setNumThreads(int numThreads)

Set the number of threads to use.

Specified by:

setNumThreads in interface TrainFlatNetwork

Parameters:

numThreads - The number of threads to use.

updateWeight

public abstract double updateWeight(double[] gradients,
                                    double[] lastGradient,
                                    int index)

Update a weight, the means by which weights are updated vary depending on the training.

Parameters:

gradients - The gradients.

lastGradient - The last gradients.

index - The index.

Returns:

The update value.

iteration

public void iteration(int count)

Perform the specified number of training iterations. This is a basic implementation that just calls iteration the specified number of times. However, some training methods, particularly with the GPU, benefit greatly by calling with higher numbers than 1.

Specified by:

iteration in interface TrainFlatNetwork

Parameters:

count - The number of training iterations.

getIteration

public int getIteration()

Specified by:

getIteration in interface TrainFlatNetwork

Returns:

The current iteration.

setIteration

public void setIteration(int iteration)

Set the iteration.

Specified by:

setIteration in interface TrainFlatNetwork

Parameters:

iteration - The iteration.