ProloGA: A Prolog implementation of a genetic algorithm

Abstract

This paper describes ProloGA, a Prolog implementation of a genetic algorithm. Chromosomes and associated parameters were stored in a Prolog database. The genetic operators of crossover, mutation, and population fitness were encoded in Prolog clauses. The test application demonstrated the feasibility of developing genetic algorithms in Prolog. The advantages of Prolog over conventional languages include database functionality, built-in 'don't care' operator, compact, declarative code, and use of heuristic knowledge. It is suggested that genetic algorithms may enhance Prolog applications by adding flexibility and adaptive rule discovery to the heuristic knowledge approach of Prolog. The combination may prove to be synergistic when applied to combinatorially large, complex, fuzzy problems.