EVOLUTION STOCHASTIC OPTIMIZATION ALGORITHMS: GENETIC ALGORITHMS AND EVOLUTION STRATEGIES

Vladimír Olej

   The applicability of multiparameter systems is determined by their optimization for the desired task. New methods of multiparameter system optimization have been published recently. They are, for example, neural networks and evolution stochastic optimization algorithms. Design and analysis of evolution stochastic optimization algorithms requires unambiguous formulating and possibility to predict the algorithm behaviour while designing them. The paper presents design and analysis of genetic algorithms (evolution strategies), new genetic algorithms (evolution strategies) with a distributed genotype, and new distributed genetic algorithms (evolution strategies). These algorithms can be used (in addition to other fields of artificial intelligence) for rectangle packing applications.

Keywords: genetic algorithms, evolution strategies, genetic algorithms and evolution strategies with a distributed genotype, distributed genetic algorithms and evolution strategies, rectangle packing applications

[full-paper]