Parallel Optimization of Evolutionary Algorithms

. A parallel two-level evolutionary algorithm which evolves genetic algorithms of maximum convergence velocity is presented. The meta-algorithm combines principles of evolution strategies and genetic algorithms in order to optimize continuous and discrete parameters of the genetic algorithms at the same time (mixed-integer optimization). The genetic algorithms which result from the meta-evolution experiment are considerably faster than standard genetic algorithms and confirm recent theoretical results about optimal mutation rates and the interaction of selective pressure and mutation rate. 1 Introduction The problem to find optimal settings of the exogenous parameters of a genetic algorithm --- with particular emphasis on the mutation rate pm , crossover probability p c , and population size --- constitutes an optimization problem by itself. From a theoretical point of view, only very few results are known which may help to find a useful parameter setting (e.g., a mutation rate pm =..