org.neuroph.contrib.bpbench

Class AbstractTraining

java.lang.Object

org.neuroph.contrib.bpbench.AbstractTraining

Direct Known Subclasses:

BackpropagationTraining, MomentumTraining, QuickpropagationTraining, ResilientTraining

public abstract class AbstractTraining
extends java.lang.Object

Abstract class that defines needed methods for benchmarking

Constructor Summary

Constructors

Constructor and Description
AbstractTraining(DataSet dataset, TrainingSettings settings) Create instance of training with new neural network
AbstractTraining(NeuralNetwork neuralNet, DataSet dataset, TrainingSettings settings) Create instance of training with given parameters

Method Summary

Modifier and Type	Method and Description
ConfusionMatrix	createMatrix() Method that creates confusion matrix from given dataset and neuraln network
DataSet	getDataset() Returns dataset
NeuralNetwork	getNeuralNet() Returns neural network
TrainingSettings	getSettings() Returns settings for training
TrainingStatistics	getStats() Returns statistics generated by this training
abstract LearningRule	setParameters() Abstract method for setting up learning rule
void	setSettings(TrainingSettings settings) Sets up settings for training
abstract void	testNeuralNet() Abstract method for executing training

Methods inherited from class java.lang.Object

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

Constructor Detail

AbstractTraining

public AbstractTraining(NeuralNetwork neuralNet,
                        DataSet dataset,
                        TrainingSettings settings)

Create instance of training with given parameters

Parameters:

neuralNet -

dataset -

settings -

AbstractTraining

public AbstractTraining(DataSet dataset,
                        TrainingSettings settings)

Create instance of training with new neural network

Parameters:

dataset -

settings -

Method Detail

testNeuralNet

public abstract void testNeuralNet()

Abstract method for executing training

setParameters

public abstract LearningRule setParameters()

Abstract method for setting up learning rule

Returns:

returns instance of learning rule with parameters

createMatrix

public ConfusionMatrix createMatrix()

Method that creates confusion matrix from given dataset and neuraln network

Returns:

confusion matrix for this training

getSettings

public TrainingSettings getSettings()

Returns settings for training

Returns:

settings for training

setSettings

public void setSettings(TrainingSettings settings)

Sets up settings for training

Parameters:

settings -

getDataset

public DataSet getDataset()

Returns dataset

Returns:

dataset used for training

getNeuralNet

public NeuralNetwork getNeuralNet()

Returns neural network

Returns:

neural network learned in this training

getStats

public TrainingStatistics getStats()

Returns statistics generated by this training

Returns:

statistics