Multi Expression Programming for solving classification problems

TL;DR

This paper presents Multi Expression Programming (MEP), a genetic programming variant that encodes multiple solutions per chromosome, demonstrating competitive performance on classification tasks compared to neural networks and linear genetic programming.

Contribution

It introduces and details strategies for applying MEP to binary and multi-class classification problems within a multi-solutions paradigm.

Findings

MEP performs similarly or better than neural networks.

MEP outperforms linear genetic programming.

Extensive experiments validate MEP's effectiveness.

Abstract

Multi Expression Programming (MEP) is a Genetic Programming variant which encodes multiple solutions in a single chromosome. This paper introduces and deeply describes several strategies for solving binary and multi-class classification problems within the \textit{multi solutions per chromosome} paradigm of MEP. Extensive experiments on various classification problems are performed. MEP shows similar or better performances than other methods used for comparison (namely Artificial Neural Networks and Linear Genetic Programming).