Fitness Landscapes and Evolutionary Algorithms
- 352 Downloads
Evolutionary algorithms (EAs) have been increasingly, and successfully, applied to combinatorial optimization problems. However, EAs are relatively complicated algorithms (compared to local search, for example) and it is not always clear to what extent their behaviour can be explained by the particular set of strategies and parameters used.
One of the most commonly-used metaphors to describe the process of simple methods such as local search is that of a ‘fitness landscape’, but even in this case, describing what we mean by such a term is not as easy as might be assumed.
In this paper, we first present some intuitive ideas and mathematical definitions of what is meant by a landscape and its properties, and review some of the theoretical and experimental work that has been carried out over the past 6 years. We then consider how the concepts associated with a landscape can be extended to search by means of evolutionary algorithms, and connect this with previous work on epistasis variance measurement.
The example of the landscapes of the Onemax function will be considered in some detail, and finally, some conclusions will be drawn on how knowledge of typical landscape properties can be used to improve the efficiency and effectiveness of heuristic search techniques.
KeywordsLocal Search Evolutionary Algorithm Travel Salesman Problem Fitness Landscape Gray Code
Unable to display preview. Download preview PDF.
- 1.Höhn, C., Reeves, C.R.: The crossover landscape for the onemax problem. In: Alander, J. (ed.) Proceedings of the 2nd Nordic Workshop on Genetic Algorithms and their Applications, pp. 27–43. University of Vaasa Press, Vaasa (1996)Google Scholar
- 2.Jones, T.C.: Evolutionary Algorithms, Fitness Landscapes and Search, Doctoral dissertation, University of New Mexico, Albuquerque, NM (1995)Google Scholar
- 5.Reeves, C.R.: Predictive measures for problem diffculty. In: Proceedings of 1999 Congress on Evolutionary Computation, pp. 736–743. IEEE Press, Los Alamitos (1999)Google Scholar
- 7.Jones, T.C.: Crossover, macromutation and population-based search. In: Eshelman, L.J. (ed.) Proceedings of the 6th International Conference on Genetic Algorithms, pp. 73–80. Morgan Kaufmann, San Francisco (1995)Google Scholar
- 8.Kauffman, S.: The Origins of Order: Self-Organization and Selection in Evolution. Oxford University Press, Oxford (1993)Google Scholar
- 16.Reeves, C.R., Yamada, T.: Embedded Path Tracing and Neighbourhood Search Techniques in Genetic Algorithms. In: Miettinen, K., Mákelä, M.M., Neittaanmäki, P., Périaux, J. (eds.) Evolutionary Algorithms in Engineering and Computer Science, pp. 95–111. John Wiley & Sons, Chichester (1999)Google Scholar
- 17.Reeves, C.R., Yamada, T.: Goal-Oriented Path Tracing Methods. In: Corne, D.A., Dorigo, M., Glover, F. (eds.) New Methods in Optimization. McGraw-Hill, London (1999)Google Scholar
- 19.Yamada, T., Reeves, C.R.: Solving the Csum permutation flowshop scheduling problem by genetic local search. In: Proc. of 1998 International Conference on Evolutionary Computation, pp. 230–234. IEEE Press, Los Alamitos (1998)Google Scholar
- 20.Reeves, C.R.: Genetic algorithms and neighbourhood search. In: Fogarty, T.C. (ed.) Evolutionary Computing: AISB Workshop, Leeds, UK, pp. 115–130. Springer, Berlin (1994) (selected Papers)Google Scholar