A perspective on the foundation and evolution of the linkage learning genetic algorithms

Intelligent guessing plays a critical role in the success and scalability o f a non-enumerative optimization algorithm that primarily relies on the sam ples taken from the search space to guide the optimization process. Linkage learning deals with the issue of intelligent guessing by exploiting proper ties of the representation. This paper underscores the importance of linkag e learning in genetic algorithms and other adaptive sampling-based optimiza tion algorithms. It develops the Foundation, identifies the problems of imp licit linkage learning in simple genetic algorithms, reviews some of the ea rly linkage learning efforts, reports some of the recent developments, and identifies the future directions of linkage learning research. (C) 2000 Els evier Science S.A. All rights reserved.