engleski

An Analysis of Recombination Techniques in Evolution Programs

The existence of overly many parameters and choices in directing the optimization (search) process based on evolutionary algorithms require additional heuristics. These heuristics may be derived from focused experiments that reveal the impact of particular parameter on the efficiency of search for a global optimum. The paper concentrates on evolution strategies (ES), a subset of evolution programs developed as methods for numerical optimization, and in particular on multimembered (lambda+mu)-ES, where lambda individuals produce mu offspring. The paper analyzes the influence of various recombination techniques on the convergence rate of (lambda+mu)-ES. Experimental explorations of various recombination techniques are performed on the optimization task of nine standard, well-known objective functions of two variables.

evolution programs ; genetic algorithms ; recombination

nije evidentirano