Genetic Programming

Volume 5, Issue 3
Summer 1995

This article presents a method for optimizing expressions to solve a given problem using the strategy of Darwinism. In contrast to genetic algorithms, which evolve an encoded representation of the solution, genetic programming evolves the solution expression directly. The Mathematica implementation makes use of the built-in features of functional programming, recursion, and hierarchical data structures. An application to symbolic regression is presented.