Fine Local Tuning

Abstract

Genetic algorithms display inherent difficulties in performing local search for numerical applications. Holland suggested [188] that the genetic algorithm should be used as a preprocessor to perform the initial search, before turning the search process over to a system that can employ domain knowledge to guide the local search. As observed in [170]:

“Like natural genetic systems, GAs progress by virtue of changing the distribution of high performance substructures in the overall population; individual structures are not the focus of attention. Once the high performance regions of the search space are identified by a GA, it may be useful to invoke a local search routine to optimize the members of the final population.”

Weeks later, when a visitor asked him what he taught his disciples, he said, ‘To get their priorities right: Better have the money than calculate it; better have the experience than define it.’

Anthony de Mello, One Minute Wisdom