A comparative study of GP-based and state-of-the-art classifiers on a synthetic machine learning benchmark

Patryk Orzechowski, Pawel Renc, William G. La Cava, Jason H. Moore, Arkadiusz Sitek, Jaroslaw Was, Joost Wagenaar

Abstract: In this paper we compare performance of genetic programming-based symbolic classifiers on a novel synthetic machine learning benchmark called DIGEN. This framework and collection of 40 different classification problems was designed specifically to differentiate performance of leading machine learning methods.