com.github.chen0040.mlp.ann

Class MLP

java.lang.Object

com.github.chen0040.mlp.ann.MLPNet

com.github.chen0040.mlp.ann.MLP

All Implemented Interfaces:

Cloneable

Direct Known Subclasses:

MLPWithLabelOutput, MLPWithNumericOutput

public abstract class MLP
extends MLPNet

Created by xschen on 21/8/15.

Field Summary

Fields inherited from class com.github.chen0040.mlp.ann.MLPNet

hiddenLayers, inputLayer, learningRate, outputLayer

Constructor Summary

Constructors

Constructor and Description
MLP()

Method Summary

Modifier and Type	Method and Description
void	copy(MLPNet rhs)
abstract double[]	getTarget(com.github.chen0040.data.frame.DataRow tuple)
protected abstract boolean	isValidTrainingSample(com.github.chen0040.data.frame.DataRow tuple)
void	setNormalizeOutputs(boolean normalize)
void	train(com.github.chen0040.data.frame.DataFrame batch, int training_epoches)
double[]	transform(com.github.chen0040.data.frame.DataRow tuple)

Methods inherited from class com.github.chen0040.mlp.ann.MLPNet

addHiddenLayer, addHiddenLayer, createInputLayer, createOutputLayer, get_target_error, minus, train, transform

Methods inherited from class java.lang.Object

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

Constructor Detail

MLP

public MLP()

Method Detail

copy

public void copy(MLPNet rhs)
          throws CloneNotSupportedException

Overrides:

copy in class MLPNet

Throws:

CloneNotSupportedException

isValidTrainingSample

protected abstract boolean isValidTrainingSample(com.github.chen0040.data.frame.DataRow tuple)

setNormalizeOutputs

public void setNormalizeOutputs(boolean normalize)

getTarget

public abstract double[] getTarget(com.github.chen0040.data.frame.DataRow tuple)

train

public void train(com.github.chen0040.data.frame.DataFrame batch,
                  int training_epoches)

transform

public double[] transform(com.github.chen0040.data.frame.DataRow tuple)