Hierarchical Modified RPSO based technique for optimal rule extraction

Author

Mukhopadhyay, Sumitra ; Mandal, Ajit K.

Author_Institution

Inst. of Radio Phys. & Electron., Univ. of Calcutta, Kolkata, India

fYear

2009

fDate

14-16 Dec. 2009

Firstpage

1

Lastpage

4

Abstract

In this paper, we present a Modified Robust Particle Swarm Optimization based learning technique for automatic extraction of fuzzy rules and subsequently for updating the parameters of a self-organized neuro-fuzzy network. The learning algorithm of network parameters is based on assigning balanced importance on local and global information. Experiments, conducted with standard benchmark problems, show the effectiveness of the method with a small number of rules along with comparable estimation error.

Keywords

fuzzy set theory; network parameters; neural nets; particle swarm optimisation; comparable estimation error; fuzzy rules automatic extraction; global information; hierarchical modified RPSO based technique; modified robust particle swarm optimization; network parameters network algorithm; neurofuzzy network; optimal rule extraction; standard benchmark problems; Cognition; Data mining; Estimation error; Function approximation; Fuzzy logic; Fuzzy neural networks; Fuzzy systems; Neural networks; Particle swarm optimization; Robustness; Accommodation Boundary; Expert System; Modified RPSO; Robust Particle Swarm Optimization;

fLanguage

English

Publisher

ieee

Conference_Titel

Computers and Devices for Communication, 2009. CODEC 2009. 4th International Conference on

Conference_Location

Kolkata

Print_ISBN

978-1-4244-5073-2

Type

conf

Filename

5407088