Information Dimension of a Population’s Attractor a Binary Genetic Algorithm
The tools for a description of chaotic dynamics are applied to investigate the work of a binary genetic algorithm (BGA). The method for determining strange attractors from BGA’s populations is shown. Attractor’s information dimension is taken as a measure of the state of BGA’s activity. The equivalence between the information dimension of a population’s attractor and the entropy of related bit positions for a given population is shown and confirmed experimentally.
KeywordsGenetic Algorithm Search Space Strange Attractor Information Dimension Poincare Section
Unable to display preview. Download preview PDF.
- Bäck T,: “The Interaction of Mutation Rate, Selection, and Self-Adaptation within a Genetic Algorithm”, Parallel Problem Solving from Nature, Vol. 2, pp. 85–94, Amsterdam, North Holland, 1992.Google Scholar
- Falconer K,J,: “Fractal geometry”, Math. Found. Appl., John Wiley, Chichester, pp.15–25, 1990.Google Scholar
- Kieś P.: “Dimension of attractors generated by a genetic algorithm”, Procs. of the IX-th Intern. Symposium on Intelligent Information Systems, pp. 40–45, Bystra, Poland, June 2000.Google Scholar
- Kieś P.: “Meaning of the information dimension of a population’s attractor in a binary GA”, IEEE Trans. on Evolutionary Computation, 2001 (sent to print).Google Scholar
- Ott E.: Chaos in Dynamical Systems, Cambridge Univ. Press, Cambridge, 1996.Google Scholar
- Qi T., Palmieri F.: “Theoretical Analysis of Evolutionary Algorithms with an Infinite Population Size in Continuous Space, Part II: Analysis of the Diversification Role of Crossover”, IEEE Trans. of Neural Networks, Vol. 5, No. 1, 1994Google Scholar
- Rudolph G.: “Convergence analysis of canonical genetic algorithms”, IEEE Trans. on Neural Networks, Special Issue on Evolutionary Computation, Vol. 5, No, 1, 1994Google Scholar