Title: Individual scatter partition-based fuzzy neural networks using particle swarm optimisation

Authors: Keon-Jun Park; Yong-Kab Kim; Byun-Gon Kim; Kwan-Woong Kim

Addresses: Department of Information and Communication Engineering, Wonkwang University, 460, Iksandae-ro, Iksan, Jeonbuk 570-749, South Korea ' Department of Information and Communication Engineering, Wonkwang University, 460, Iksandae-ro, Iksan, Jeonbuk 570-749, South Korea ' Department of Electronic Engineering, Kunsan National University, 1170, Daehangno, Gunsan 573-701, South Korea ' Director in Digital Signal Processing Team, Thunder Technology, 1766, Yonjang-Li, Jinan-Yup, Jinan-Gun, Jeonbuk 561-221, South Korea

Abstract: This paper presents a new design of fuzzy neural networks (FNNs) based on individual scatter partition using particle swarm optimisation (PSO). The proposed FNNs are expressed by the scatter partition of input space generated by fuzzy c-means clustering algorithm. The partitioned local spaces indicate the fuzzy rules of the FNNs that have the individual regions in the different size. The consequence part of the rule is represented by polynomial functions. The back propagation algorithm is used to estimate the coefficients of the polynomial functions. The optimisation to find individual regions and parameters of learning is conducted by PSO. The performance of the proposed FNNs is demonstrated with the non-linear process.

Keywords: FNNs; fuzzy neural networks; individual scatter partition; fuzzy c-means; clustering algorithms; PSO; particle swarm optimisation; nonlinear processes; polynomial functions.

DOI: 10.1504/IJSNET.2014.064433

International Journal of Sensor Networks, 2014 Vol.15 No.4, pp.223 - 230

Received: 11 Feb 2014
Accepted: 16 Feb 2014
Published online: 25 Aug 2014 *

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