Self-adaptive Operator Scheduling Using the Religion-Based EA
The optimal choice of the variation operators mutation and crossover and their parameters can be decisive for the performance of evolutionary algorithms (EAs). Usually the type of the operators (such as Gaussian mutation) remains the same during the entire run and the probabilistic frequency of their application is determined by a constant parameter, such as a fixed mutation rate. However, recent studies have shown that the optimal usage of a variation operator changes during the EA run. In this study, we combined the idea of self-adaptive mutation operator scheduling with the Religion-Based EA (RBEA), which is an agent model with spatially structured and variable sized subpopulations (religions). In our new model (OSRBEA), we used a selection of different operators, such that each operator type was applied within one specific subpopulation only. Our results indicate that the optimal choice of operators is problem dependent, varies during the run, and can be handled by our self-adaptive OSRBEA approach. Operator scheduling could clearly improve the performance of the already very powerful RBEA and was superior compared to a classic and other advanced EA approaches.
Unable to display preview. Download preview PDF.
- 1.Krink, T. and Ursem, R.K.: Parameter Control Using the Agent Based Patchwork Model. In: Proceedings of the Second Congress on Evolutionary Computation (CEC-2000), Vol. 1. San Diego, CA, USA (2000) 77–83Google Scholar
- 2.Bäck, T.: Self-Adaptation in Genetic Algorithms. In: F. J. Varela, P.B. (ed.): Proceedings of 1st European Conference on Artificial Life. MIT Press (1992) 263–271Google Scholar
- 3.Eiben, A.E., Sprinkhuizen-Kuyper, I.G., and Thijssen, B.A.: Competing Crossovers in an Adaptive GA Framework. In: Proceedings of the 1998 IEEE International Conference on Evolutionary Computation (1998) 787–792Google Scholar
- 4.Smith, J.E. and Fogarty, T.C.: Self Adaptation of Mutation Rates in a Steady State Genetic Algorithm. In: Proceedings of the 1996 IEEE International Conference on Evolutionary Computing, IEEE Press (1996) 318–323Google Scholar
- 5.Davis, L.: Adapting Operator Probabilities in Genetic Algorithms. In: Schaffer, J.D. (ed.): Proceedings of the Third International Conference on Genetic Algorithms (ICGA III) (1989) 61–69Google Scholar
- 6.Schlierkamp-Voosen, D. and Mühlenbein, H.: Strategy Adaptation by Competing Subpopulations. In: Davidor, Y., Schwefel, H.P., and Maenner, R. (eds.): Proceedings of the Third International Conference on Parallel Problem Solving from Nature (PPSN III) (1994) 199–208Google Scholar
- 7.Thomsen, R., Rickers, P., and Krink, T.: A Religion-Based Spatial Model For Evolutionary Algorithms. In: Schoenauer, M., Deb, K., Rudolph, G., Yao, X., Lutton, E., Merelo, J.J., and Schwefel, H.P. (eds.): Parallel Problem Solving from Nature (PPSN VI) (2000) 817–826Google Scholar
- 8.Whitley, D.: Cellular Genetic Algorithms. In: Forrest, S. (ed.): Proceedings of the Fifth International Conference on Genetic Algorithms, Morgan Kaufman (1993) 658Google Scholar
- 9.Bäck, T., Fogel, D.B., Michalewicz, Z., et al. (eds.).: Handbook on Evolutionary Computation. IOP Publishing Ltd and Oxford University Press (1997).Google Scholar
- 10.Krink, T., Rickers, P., and Thomsen, R.: Applying Self-Organised Criticality to Evolutionary Algorithms. In: Schoenauer, M., Deb, K., Rudolph, G., Yao, X., Lutton, E., Merelo, J.J., and Schwefel, H.P. (eds.): Parallel Problem Solving from Nature (PPSN VI) (2000) 375–384Google Scholar
- 11.Thomsen, R. and Rickers, P. Introducing Spatial Agent-Based Models and Self-Organised Criticality to Evolutionary Algorithms. Master’s thesis, University of Aarhus, Denmark (2000)Google Scholar