Meta-optimization

[jenetics]

In general, a genetic algorithm is used to find input parameters, which optimizes a given fitness function. To do this you have to create an instance of an evolution Engine, which itself has to be parameterized. The engine parameters, which influences the evolution performance are defined in the EvolutionParams class.

public final class EvolutionParams {
    final Selector<G, C> survivorsSelector;
    final Selector<G, C> offspringSelector;
    final Alterer<G, C> alterer;
    final int populationSize;
    final double offspringFraction;
    final long maximalPhenotypeAge;
}

The purpose of this ticket is to support the optimization of the evolution parameters for a given optimization problem. The assumption is, that for a given problem class there exists an optimal set of evolution parameters.