A new evolutionary system for evolving artificial neural networks

TLDR

This paper introduces EPNet, an evolutionary system that emphasizes evolving ANN behaviors rather than merely their structure. EPNet employs Fogel’s evolutionary programming with five behavior‑focused mutation operators, simultaneously evolving network architecture and weights while preserving behavioral links through partial training and node splitting, and favoring deletion to promote parsimony. On benchmark problems such as parity, medical diagnosis, credit card assessment, and Mackey‑Glass prediction, EPNet produced compact networks with strong generalization, outperforming other algorithms.

Abstract

This paper presents a new evolutionary system, i.e., EPNet, for evolving artificial neural networks (ANNs). The evolutionary algorithm used in EPNet is based on Fogel's evolutionary programming (EP). Unlike most previous studies on evolving ANN's, this paper puts its emphasis on evolving ANN's behaviors. Five mutation operators proposed in EPNet reflect such an emphasis on evolving behaviors. Close behavioral links between parents and their offspring are maintained by various mutations, such as partial training and node splitting. EPNet evolves ANN's architectures and connection weights (including biases) simultaneously in order to reduce the noise in fitness evaluation. The parsimony of evolved ANN's is encouraged by preferring node/connection deletion to addition. EPNet has been tested on a number of benchmark problems in machine learning and ANNs, such as the parity problem, the medical diagnosis problems, the Australian credit card assessment problem, and the Mackey-Glass time series prediction problem. The experimental results show that EPNet can produce very compact ANNs with good generalization ability in comparison with other algorithms.