A genetics-based technique for the automated acquisition of expert system rule bases

Abstract

The genetic algorithm (GA) is a powerful search paradigm which combines elements from evolutionary biology with concepts from population genetics. Because they operate in a domain-independent fashion, GAs have been successfully applied to a wide variety of optimization and learning problems. A technique is presented by which genetic algorithms can be adapted to operate upon the LISP-like production rules typically used in expert systems. A brief overview is presented of genetic algorithms and genetics-based learning.<>