private static IBitMutation <IBitIndividual> CreateAlgo(Func <IBitMutation <IBitIndividual> > mutation,
int geneCount)
{
var m = mutation.Invoke();
var algo = new BitEvolutionaryAlgorithm <IBitIndividual>()
.UsingBasicStatistics()
.UsingRandomPopulation()
.UsingParameters(new Parameters
{
Mu = 1,
Lambda = 2,
GeneCount = geneCount,
MutationRate = 2
}).UsingEvaluation(new OneMaxFitness <IBitIndividual>())
.UsingHeuristic(
e => e
.CloneGenesFrom(new FirstParentSelector <IBitIndividual>())
.ThenApplyMutation(m),
new BitElitismGenerationFilter <IBitIndividual>(true));
algo.Initialize();
return(m);
}
public void Func()
{
const int geneCount = 500;
const int mutationRate = 2;
const int repairRate = 2;
const int learningRate = 2;
const int observationPhase = 10;
const double beta = 1.5;
var endogenous = new BitEvolutionaryAlgorithm <IBitIndividual>()
.UsingParameters(new Parameters
{
GeneCount = geneCount,
// Self adapting
MutationRate = mutationRate,
// Constant, based on gene count
Lambda = (int)(5 * Math.Log(geneCount)),
// Always 1
Mu = 1,
})
.UsingBasicStatistics()
.UsingRandomPopulation(mutationRate)
.UsingHeuristic(SingleEndogenous(learningRate))
.UsingEvaluation(new OneMaxFitness <IBitIndividual>());
var stagnation = new BitEvolutionaryAlgorithm <IBitIndividual>()
.UsingParameters(new Parameters
{
GeneCount = geneCount,
// Self adapting
MutationRate = mutationRate,
// Constant, based on gene count
Lambda = (int)(3 * Math.Log(geneCount)),
// Always 1
Mu = 1,
})
.UsingStagnationStatistics()
.UsingRandomPopulation(learningRate)
.UsingHeuristic(StagnationDetection(learningRate, mutationRate, 1))
.UsingEvaluation(new OneMaxFitness <IBitIndividual>());
var asymmetric = new BitEvolutionaryAlgorithm <IBitIndividual>()
.UsingParameters(new Parameters
{
GeneCount = geneCount,
// Self adapting, based on learning rate
MutationRate = 3,
// Always 1
Lambda = 1,
// Always 1
Mu = 1,
})
.UsingRandomPopulation(mutationRate)
.UsingStatistics(new AsymmetricBasicEvolutionaryStatistics <IBitIndividual>())
.UsingHeuristic(Asymmetric(0.05, observationPhase))
.UsingEvaluation(new OneMaxFitness <IBitIndividual>());
var oneLambdaLambda = new BitEvolutionaryAlgorithm <IBitIndividual>()
.UsingParameters(new Parameters
{
GeneCount = geneCount,
// Constant
MutationRate = mutationRate,
// Self-adapting, initial 1
Lambda = 1,
// Always 1
Mu = 1,
})
.UsingStatistics(new LambdaBasicStatistics <IBitIndividual>())
.UsingRandomPopulation(mutationRate)
.UsingHeuristic(Repair(learningRate, repairRate))
.UsingEvaluation(new OneMaxFitness <IBitIndividual>());
var heavyTail = new BitEvolutionaryAlgorithm <IBitIndividual>()
.UsingParameters(new Parameters
{
GeneCount = geneCount,
// Self adapting
MutationRate = mutationRate,
// Constant
Lambda = (int)(3 * Math.Log(geneCount)),
// Always 1
Mu = 1,
})
.UsingBasicStatistics()
.UsingRandomPopulation(mutationRate)
.UsingHeuristic(HeavyTail(learningRate, beta))
.UsingEvaluation(new OneMaxFitness <IBitIndividual>());
var lambdaEndogenous = new BitEvolutionaryAlgorithm <IBitIndividual>()
.UsingParameters(new Parameters
{
GeneCount = geneCount,
// Self adapting
MutationRate = mutationRate,
// Constant, based on gene count
Lambda = (int)(3 * Math.Log(geneCount)),
// Constant, based on gene count
Mu = (int)(3 * Math.Log(geneCount)),
})
.UsingStatistics(new EndogenousBasicEvolutionaryStatistics())
.UsingRandomPopulation(mutationRate)
.UsingHeuristic(MultiEndogenous(learningRate))
.UsingEvaluation(new OneMaxFitness <IBitIndividual>());
}
