Adaptive Learning by Genetic Algorithms

I started to deal with genetic algorithms in 1993 when I was working on a project on learning and rational behavior in economic systems. Initially I carried out simulations in an overlapping generations model but soon got dissatisfied with the complete lack of theoretical foundation for the observed behavior. Thus, I started to work on a mathematical representation of the behavior of a simple genetic algorithm in the special setup of an interacting population of economic agents and step by step arrived at the results collected here. However, I believe that much more can and has to be done in this field. I would like to thank Gustav Feichtinger who not only supervised my doctoral thesis but always supported and encouraged me throughout the last few years. Special thanks are also due to K. Hornik, A. Mehlmann and M. Kopel who contributed largely to the work. During the preparation of the monograph I also benefited from helpful comments of A. Geyer-Schulz, G. Rote, G. Tragler and A. Rahman. Special thanks to W. A. Muller from Springer-Verlag for his support. Financial support from the Austrian Science Foundation under contract number P9112-S0Z is gratefully acknowledged.