GAFeedForwardNN.h

libcudann

* GAFeedForwardNN.h

* Created on: Jan 20, 2011

* Author: donati

#ifndef GAFEEDFORWARDNN_H_

#define GAFEEDFORWARDNN_H_

#include "FeedForwardNN.h"

#include "LearningSet.h"

#include "FloatChromosome.h"

#include "FeedForwardNNTrainer.h"

#define ROULETTE_WHEEL 0

#define TOURNAMENT_SELECTION 1

//performs crossover between a chromosome and a mate, and return the result

FloatChromosome crossover(const FloatChromosome & first, const FloatChromosome & second, const float pcross);

//performs mutation on a chromosome, and return the result

FloatChromosome mutation(const FloatChromosome & first, const float pmut, const int maxdimhiddenlayers, const int nhidlayers);

//roulette wheel selection

int rouletteWheel(const int size, const float * fitnesses);

//tournament selection

int tournament(const int size, const float * fitnesses, const int toursize);

class GAFeedForwardNN {

public:

GAFeedForwardNN();

virtual ~GAFeedForwardNN();

//choose the training set

void selectTrainingSet(LearningSet &);

//choose the test set. if this is set the error rate is computed on test set instead of training set

void selectTestSet(LearningSet &);

//choose a net to save the best network individual trained so far. mse on test set (or train if no test is specified) is the criterion

void selectBestNet(FeedForwardNN & n);

//initialize the genetic algorithm parameters and create the first population of individuals

void init(const int popsize,const int generations,const int selectionalgorithm,const int numberofevaluations,const float pcross,const float pmut,const int nhidlayers,const int maxdimhiddenlayers);

//run the genetic algorithm initialized before with some training parameters:

//training location, training algorithm, desired error, max_epochs, epochs_between_reports

//see "FeedForwardNNTrainer" class for more details

//printtype specifies how much verbose will be the execution (PRINT_ALL,PRINT_MIN,PRINT_OFF)

void evolve(const int n, const float * params, const int printtype=PRINT_ALL);

private:

FeedForwardNN * net;

LearningSet * trainingSet;

LearningSet * testSet;

FeedForwardNN * bestNet;

FloatChromosome * chromosomes;

int popsize;

int generations;

int numberofevaluations;

int selectionalgorithm;

float pcross;

float pmut;

int nhidlayers;

int maxdimhiddenlayers;

};

#endif /* GAFEEDFORWARDNN_H_ */