COGIN: Symbolic Induction Using Genetic Algorithms

July 1992

COGIN: Symbolic Induction Using Genetic Algorithms

Authors:

D. P. Greene and Stephen Smith

Abstract:

COGIN is a system designed for induction of symbolic decision models from pre-classed examples based on the use of genetic algorithms (GAS). Much research in symbolic induction has focused on techniques for reducing classification inaccuracies that arise from inherent limits of underlying incremental search techniques. Genetic Algorithms offer an intriguing alternative to step-wise model construction, relying instead on model evolution through global competition. The difficulty is in providing an effective framework for the GA to be practically applied to complex induction problems. COGIN merges traditional induction concepts with genetic search to provide such a framework, and recent experimental results have demonstrated its advantage relative to basic stepwise inductive approaches. In this paper, we describe the essential elements of the COGIN approach and present a favorable comparison of COGIN results with those produced by a more sophisticated stepwise approach (with support post processing) on standardized multiplexor problems.
@conference{Greene-1992-13388,
author = {D. P. Greene And Stephen Smith},
title = {COGIN: Symbolic Induction Using Genetic Algorithms},
booktitle = {Proceedings of 10th National Conference on Artificial Intelligence (AAAI '92)},
year = {1992},
month = {July},
pages = {111 - 116},
}
Copyright notice: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. These works may not be reposted without the explicit permission of the copyright holder.