conscience_core::ai::nn::CscNeuralNetworkGeneticAlgorithmGenerationGenerator Class Reference

Generates generations for a neural network genetic algorithm. More...

#include <CscNeuralNetworkGeneticAlgorithmGenerationGenerator.h>

Classes
struct	NeuralNetworkGeneticAlgorithmGeneration
	Represents a generation in the neural network genetic algorithm. More...

Public Member Functions
	CscNeuralNetworkGeneticAlgorithmGenerationGenerator (unsigned populationSizeInit, unsigned nbNeuralNetworkInputsInit, unsigned nbNeuralNetworkOutputsInit, unsigned copyPercentOfChanceInit, unsigned crossoverPercentOfChanceInit, unsigned mutationPercentOfChanceInit, CscNeuralNetworkGeneticAlgorithmIndividualGenerator::AllPossibilitiesHyperparameters *allPossibilitiesHyperparameters)
virtual	~CscNeuralNetworkGeneticAlgorithmGenerationGenerator ()
CscNeuralNetworkGeneticAlgorithmIndividualGenerator *	getIndividualGenerator () const
NeuralNetworkGeneticAlgorithmGeneration *	getPreviousGeneration () const
NeuralNetworkGeneticAlgorithmGeneration *	getCurrentGeneration () const
const unsigned	getPopulationSize () const
const unsigned	getCopyPercentOfChance () const
const unsigned	getCrossoverPercentOfChance () const
const unsigned	getMutationPercentOfChance () const
CscNeuralNetworkGeneticAlgorithmIndividualGenerator::HyperparametersIndividual *	selectParentByTournament (const unsigned nbParticipantsInTournament=DEFAULT_NB_PARTICIPANTS_IN_TOURNAMENT, const float probaForBestInTournament=DEFAULT_PROBA_FOR_BEST_IN_TOURNAMENT)
	Selects a parent for the next generation using the tournament selection method. More...
void	sortCurrentGenerationByResult ()
void	fillCurrentGenerationRandomly ()
void	fillNewCurrentGeneration (unsigned nthGeneration)
void	printResultsGenerationCreation ()

Detailed Description

Generates generations for a neural network genetic algorithm.

This class is responsible for managing and generating generations of a neural network genetic algorithm. It includes methods for selecting parents, sorting generations, and creating new individuals.

Constructor & Destructor Documentation

CscNeuralNetworkGeneticAlgorithmGenerationGenerator()

conscience_core::ai::nn::CscNeuralNetworkGeneticAlgorithmGenerationGenerator::CscNeuralNetworkGeneticAlgorithmGenerationGenerator	(	unsigned	populationSizeInit,
		unsigned	nbNeuralNetworkInputsInit,
		unsigned	nbNeuralNetworkOutputsInit,
		unsigned	copyPercentOfChanceInit,
		unsigned	crossoverPercentOfChanceInit,
		unsigned	mutationPercentOfChanceInit,
		CscNeuralNetworkGeneticAlgorithmIndividualGenerator::AllPossibilitiesHyperparameters *	allPossibilitiesHyperparameters
	)

~CscNeuralNetworkGeneticAlgorithmGenerationGenerator()

conscience_core::ai::nn::CscNeuralNetworkGeneticAlgorithmGenerationGenerator::~CscNeuralNetworkGeneticAlgorithmGenerationGenerator ( )

virtual

Member Function Documentation

fillCurrentGenerationRandomly()

void conscience_core::ai::nn::CscNeuralNetworkGeneticAlgorithmGenerationGenerator::fillCurrentGenerationRandomly

(

)

fillNewCurrentGeneration()

void conscience_core::ai::nn::CscNeuralNetworkGeneticAlgorithmGenerationGenerator::fillNewCurrentGeneration

(

unsigned

nthGeneration

)

getCopyPercentOfChance()

const unsigned conscience_core::ai::nn::CscNeuralNetworkGeneticAlgorithmGenerationGenerator::getCopyPercentOfChance

(

)

const

getCrossoverPercentOfChance()

const unsigned conscience_core::ai::nn::CscNeuralNetworkGeneticAlgorithmGenerationGenerator::getCrossoverPercentOfChance

(

)

const

getCurrentGeneration()

CscNeuralNetworkGeneticAlgorithmGenerationGenerator::NeuralNetworkGeneticAlgorithmGeneration * conscience_core::ai::nn::CscNeuralNetworkGeneticAlgorithmGenerationGenerator::getCurrentGeneration

(

)

const

getIndividualGenerator()

CscNeuralNetworkGeneticAlgorithmIndividualGenerator * conscience_core::ai::nn::CscNeuralNetworkGeneticAlgorithmGenerationGenerator::getIndividualGenerator

(

)

const

getMutationPercentOfChance()

const unsigned conscience_core::ai::nn::CscNeuralNetworkGeneticAlgorithmGenerationGenerator::getMutationPercentOfChance

(

)

const

getPopulationSize()

const unsigned conscience_core::ai::nn::CscNeuralNetworkGeneticAlgorithmGenerationGenerator::getPopulationSize

(

)

const

getPreviousGeneration()

CscNeuralNetworkGeneticAlgorithmGenerationGenerator::NeuralNetworkGeneticAlgorithmGeneration * conscience_core::ai::nn::CscNeuralNetworkGeneticAlgorithmGenerationGenerator::getPreviousGeneration

(

)

const

printResultsGenerationCreation()

void conscience_core::ai::nn::CscNeuralNetworkGeneticAlgorithmGenerationGenerator::printResultsGenerationCreation

(

)

selectParentByTournament()

CscNeuralNetworkGeneticAlgorithmIndividualGenerator::HyperparametersIndividual * conscience_core::ai::nn::CscNeuralNetworkGeneticAlgorithmGenerationGenerator::selectParentByTournament	(	const unsigned	nbParticipantsInTournament = DEFAULT_NB_PARTICIPANTS_IN_TOURNAMENT,
		const float	probaForBestInTournament = DEFAULT_PROBA_FOR_BEST_IN_TOURNAMENT
	)

Selects a parent for the next generation using the tournament selection method.

This method conducts a tournament selection process, which is a stochastic method used to select the best individual based on their fitness scores from a subset of the current generation. It randomly picks a specified number of participants from the previous generation, sorts them based on their performance, and then probabilistically selects one to be a parent based on their rank. The probability of selecting each participant decreases for lower-ranked individuals, ensuring that the best performers have a higher chance of being chosen, while still allowing for genetic diversity.

1. Randomly select a set number of individuals for the tournament.
2. Sort these individuals by their fitness scores in descending order.
3. Select a winner based on a probabilistic model where higher-ranked individuals are more likely to win.

Returns

Pointer to the HyperparametersIndividual that wins the tournament and will contribute to creating the new generation.

sortCurrentGenerationByResult()

void conscience_core::ai::nn::CscNeuralNetworkGeneticAlgorithmGenerationGenerator::sortCurrentGenerationByResult

(

)

---

The documentation for this class was generated from the following files:

• ArtificialIntelligence/CscNeuralNetworkGeneticAlgorithmGenerationGenerator.h
• ArtificialIntelligence/CscNeuralNetworkGeneticAlgorithmGenerationGenerator.cpp