Toward a fuzzy government of genetic populations

Although genetic algorithms (GAs) are easy to implement and are powerful tools to solve difficult problems featuring huge search spaces, they usually require human supervision to be exploited successfully. It seems that fuzzy logic techniques can help reduce the amount of human intervention needed to use GAs. The paper concentrates on a particular application to symbolic regression to illustrate how to build a fuzzy knowledge-based system, or, to use a suggestive term, a fuzzy government, for GA control.<<ETX>>