C# Class Encog.Neural.Networks.Training.SVM.SVMTrain

Provides training for Support Vector Machine networks.
Inheritance: BasicTraining
Afficher le fichier Open project: encog/encog-silverlight-core

Méthodes publiques

Méthode Description
CrossValidate ( int index, double gamma, double c ) : double

Cross validate and check the specified index/gamma.
FinishTraining ( ) : void

Called to finish training.
Iteration ( ) : void

Perform one training iteration.
SVMTrain ( BasicNetwork network, INeuralDataSet training ) : System

Construct a trainer for an SVM network.
Train ( ) : void

Quickly train all outputs with a C of 1.0 and a gamma equal to 1/(num inputs).
Train ( double gamma, double c ) : void

Quickly train the network with a fixed gamma and C.
Train ( int index, double gamma, double c ) : void

Quickly train one output with the specified gamma and C.

Private Methods

Méthode Description
Evaluate ( svm_parameter param, svm_problem prob, double target ) : double

Evaluate the error for the specified model.
Setup ( ) : void

Setup to train the SVM.

Method Details

CrossValidate() public méthode

Cross validate and check the specified index/gamma.
public CrossValidate ( int index, double gamma, double c ) : double
index int The output index to cross validate.
gamma double The gamma to check.
c double The C to check.
Résultat double

FinishTraining() public méthode

Called to finish training.
public FinishTraining ( ) : void
Résultat void

Iteration() public méthode

Perform one training iteration.
public Iteration ( ) : void
Résultat void

SVMTrain() public méthode

Construct a trainer for an SVM network.
public SVMTrain ( BasicNetwork network, INeuralDataSet training ) : System
network BasicNetwork The network to train.
training INeuralDataSet The training data for this network.
Résultat System

Train() public méthode

Quickly train all outputs with a C of 1.0 and a gamma equal to 1/(num inputs).
public Train ( ) : void
Résultat void

Train() public méthode

Quickly train the network with a fixed gamma and C.
public Train ( double gamma, double c ) : void
gamma double The gamma to use.
c double The C to use.
Résultat void

Train() public méthode

Quickly train one output with the specified gamma and C.
public Train ( int index, double gamma, double c ) : void
index int The output to train.
gamma double The gamma to train with.
c double The C to train with.
Résultat void