net.sourceforge.cilib.neuralnetwork.foundation
Class EvaluationMediator

java.lang.Object
  net.sourceforge.cilib.algorithm.Algorithm
      net.sourceforge.cilib.algorithm.SingularAlgorithm
          net.sourceforge.cilib.neuralnetwork.foundation.EvaluationMediator

All Implemented Interfaces:

Serializable, Runnable, Cloneable

Direct Known Subclasses:

FFNNEvaluationMediator, SAILAEvaluationMediator

public class EvaluationMediator
extends SingularAlgorithm

Author:

stefanv

See Also:

Serialized Form

Field Summary

protected NeuralNetworkData	data
protected NNError[]	errorDg
protected NNError[]	errorDt
protected NNError[]	errorDv
protected int	nrEvaluationsPerEpoch
protected NNError[]	prototypeError
protected NeuralNetworkTopology	topology
protected int	totalEvaluations
protected TrainingStrategy	trainer

Fields inherited from class net.sourceforge.cilib.algorithm.Algorithm

optimisationProblem

Constructor Summary

EvaluationMediator()

Method Summary

void	addPrototypError(NNError proto)
void	algorithmIteration() The actual operations that the current Algorithm performs within a single iteration.
void	computeErrorIteration(NNError[] err, TypeList output, NNPattern input)
TypeList	evaluate(NNPattern p)
void	finaliseErrors(NNError[] err)
OptimisationSolution	getBestSolution() Get the best current solution.
Algorithm	getClone() Create a cloned copy of the current object and return it.
NeuralNetworkData	getData()
EpochStrategy	getEpochStrategy()
NNError[]	getErrorDg()
NNError[]	getErrorDt()
NNError[]	getErrorDv()
int	getNrEvaluationsPerEpoch()
NNError[]	getPrototypeError()
List<OptimisationSolution>	getSolutions() Get the collection of best solutions.
NeuralNetworkTopology	getTopology()
int	getTotalEvaluations()
TrainingStrategy	getTrainer()
void	incrementEvaluationsPerEpoch()
void	performInitialisation() Perform the needed initialisation required before the execution of the algorithm starts.
void	performLearning()
void	resetError(NNError[] err)
void	setData(NeuralNetworkData data)
void	setEpochStrategy(EpochStrategy epochStrategy)
void	setErrorDg(NNError[] errorDg)
void	setErrorDt(NNError[] errorDt)
void	setErrorDv(NNError[] errorDv)
void	setErrorNoOutputs(NNError[] err, int nr)
void	setErrorNoPatterns(NNError[] err, int nr)
void	setTopology(NeuralNetworkTopology topology)
void	setTrainer(TrainingStrategy trainer)

Methods inherited from class net.sourceforge.cilib.algorithm.Algorithm

addAlgorithmListener, addStoppingCondition, get, getAlgorithmList, getIterations, getOptimisationProblem, getPercentageComplete, getStoppingConditions, initialise, isFinished, performIteration, performUninitialisation, removeAlgorithmListener, removeStoppingCondition, reset, run, setOptimisationProblem, terminate

Methods inherited from class java.lang.Object

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

Field Detail

prototypeError

protected NNError[] prototypeError

errorDg

protected NNError[] errorDg

errorDt

protected NNError[] errorDt

errorDv

protected NNError[] errorDv

nrEvaluationsPerEpoch

protected int nrEvaluationsPerEpoch

topology

protected NeuralNetworkTopology topology

data

protected NeuralNetworkData data

trainer

protected TrainingStrategy trainer

totalEvaluations

protected int totalEvaluations

Constructor Detail

EvaluationMediator

public EvaluationMediator()

Method Detail

performInitialisation

public void performInitialisation()

Description copied from class: Algorithm

Perform the needed initialisation required before the execution of the algorithm starts.

Overrides:

performInitialisation in class Algorithm

algorithmIteration

public void algorithmIteration()

Description copied from class: SingularAlgorithm

The actual operations that the current Algorithm performs within a single iteration.

Specified by:

algorithmIteration in class SingularAlgorithm

computeErrorIteration

public void computeErrorIteration(NNError[] err,
                                  TypeList output,
                                  NNPattern input)

evaluate

public TypeList evaluate(NNPattern p)

getErrorDg

public NNError[] getErrorDg()

getErrorDt

public NNError[] getErrorDt()

getErrorDv

public NNError[] getErrorDv()

getNrEvaluationsPerEpoch

public int getNrEvaluationsPerEpoch()

getTopology

public NeuralNetworkTopology getTopology()

getTotalEvaluations

public int getTotalEvaluations()

performLearning

public void performLearning()

finaliseErrors

public void finaliseErrors(NNError[] err)

resetError

public void resetError(NNError[] err)

setErrorDg

public void setErrorDg(NNError[] errorDg)

setErrorDt

public void setErrorDt(NNError[] errorDt)

setErrorDv

public void setErrorDv(NNError[] errorDv)

setErrorNoOutputs

public void setErrorNoOutputs(NNError[] err,
                              int nr)

setErrorNoPatterns

public void setErrorNoPatterns(NNError[] err,
                               int nr)

setTopology

public void setTopology(NeuralNetworkTopology topology)

getPrototypeError

public NNError[] getPrototypeError()

addPrototypError

public void addPrototypError(NNError proto)

getData

public NeuralNetworkData getData()

setData

public void setData(NeuralNetworkData data)

getTrainer

public TrainingStrategy getTrainer()

setTrainer

public void setTrainer(TrainingStrategy trainer)

getEpochStrategy

public EpochStrategy getEpochStrategy()

setEpochStrategy

public void setEpochStrategy(EpochStrategy epochStrategy)

getBestSolution

public OptimisationSolution getBestSolution()

Description copied from class: Algorithm

Get the best current solution. This best solution is determined from the personal bests of the particles.

Specified by:

getBestSolution in class Algorithm

Returns:

The OptimisationSolution representing the best solution.

getClone

public Algorithm getClone()

Description copied from class: Algorithm

Create a cloned copy of the current object and return it. In general the created copy will be a deep copy of the provided instance. As a result this operation an be quite expensive if used incorrectly.

Specified by:

getClone in interface Cloneable

Specified by:

getClone in class Algorithm

Returns:

An exact clone of the current object instance.

See Also:

Object.clone()

getSolutions

public List<OptimisationSolution> getSolutions()

Description copied from class: Algorithm

Get the collection of best solutions. This result does not actually make sense in the normal PSO algorithm, but rather in a MultiObjective optimization.

Specified by:

getSolutions in class Algorithm

Returns:

The Collection<OptimisationSolution> containing the solutions.

incrementEvaluationsPerEpoch

public void incrementEvaluationsPerEpoch()