Title: A mixed-coding adaptive differential evolution for optimising the architecture and parameters of feedforward neural networks

Authors: Li Zhang; Hong Li

Addresses: School of Mathematics and Statistics, Xidian University, Xi'an 710071, Shaanxi, China ' School of Mathematics and Statistics, Xidian University, Xi'an 710071, Shaanxi, China

Abstract: This paper presents an adaptive differential evolution with mixed-coding strategy to evolve feedforward neural networks (FNNs). This algorithm with adaptive control parameters which can handle effectively binary variables and real variables, is used to optimise simultaneously FNN architecture and connection parameters (weights and biases) by a specific individual representation and evolutionary scheme. The performance of the algorithm has been evaluated on several benchmarks. The results demonstrate that the proposed algorithm can produce compact FNNs with good generalisation ability.

Keywords: FNN; feedforward neural network; evolutionary neural network; differential evolution; generalisation ability.

DOI: 10.1504/IJSNET.2019.098556

International Journal of Sensor Networks, 2019 Vol.29 No.4, pp.262 - 274

Received: 04 Jul 2018
Accepted: 29 Jul 2018
Published online: 27 Mar 2019 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article