Package me.damoebe.architectures.mlp

Class EvolutionNetwork

java.lang.Object

me.damoebe.architectures.Network

me.damoebe.architectures.mlp.FFNetwork

me.damoebe.architectures.mlp.EvolutionNetwork

public class EvolutionNetwork
extends FFNetwork

An object class for EvolutionNetworks

Field Summary

Fields inherited from class me.damoebe.architectures.mlp.FFNetwork

currentLoss, layers, learningRate, noise

Constructor Summary

Constructors

Constructor	Description
EvolutionNetwork(int inputSize, int outputSize, int hiddenLayerSize, int hiddenLayerAmount)	Main constructor for EvolutionNetwork accessing the super constructor of FFNetwork abstract class

Method Summary

Modifier and Type	Method	Description
FFNetwork	clone()	A method for cloning a FFNetwork object
void	finishEpoch()	Finishes an epoch by saving potential success and mutating connection weights
double	getNetworkLoss()	Gets the network loss
void	train(List<Double> input, List<Double> optimalOutput)	The abstract train function
void	updateLoss(List<Double> optimalActivations)	Updates the current network loss and adds loss to epoch loss list

Methods inherited from class me.damoebe.architectures.mlp.FFNetwork

generateLayer, getHiddenLayerAmount, getHiddenLayerSize, getInputSize, getLayers, getOutput, getOutputSize, insertInput, setNoise, updateAllActivations

Methods inherited from class me.damoebe.architectures.Network

loadNetworkFromJson, loadToJsonFile

Methods inherited from class java.lang.Object

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

Constructor Details

EvolutionNetwork

public EvolutionNetwork(int inputSize,
 int outputSize,
 int hiddenLayerSize,
 int hiddenLayerAmount)

Main constructor for EvolutionNetwork accessing the super constructor of FFNetwork abstract class

Parameters:

inputSize - The size of the first layer

outputSize - The size of the last layer

hiddenLayerSize - The size of the hiddenlayers

hiddenLayerAmount - The amount of hiddenlayers that should be generated

Method Details

train

public void train(List<Double> input,
 List<Double> optimalOutput)

Description copied from class: FFNetwork

The abstract train function

Specified by:

train in class FFNetwork

Parameters:

input - The inputs for this repetition

optimalOutput - The optimalOutputs for this repetition

updateLoss

public void updateLoss(List<Double> optimalActivations)

Updates the current network loss and adds loss to epoch loss list

Overrides:

updateLoss in class FFNetwork

Parameters:

optimalActivations - The target activations for the current repetition

finishEpoch

public void finishEpoch()

Finishes an epoch by saving potential success and mutating connection weights

getNetworkLoss

public double getNetworkLoss()

Description copied from class: FFNetwork

Gets the network loss

Specified by:

getNetworkLoss in class FFNetwork

Returns:

The network loss double

clone

public FFNetwork clone()

Description copied from class: FFNetwork

A method for cloning a FFNetwork object

Specified by:

clone in class FFNetwork

Returns:

the cloned network object