Refinement of fuzzy rule weights with particle swarm optimisation

Tianhua Chen, Qiang Shen, Pan Su, Changjing Shang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

The most challenging problem in the design of fuzzy rule-based classification systems is the construction of a fuzzy rule base for the target problem. Much research has focused on generating and adjusting antecedent fuzzy sets. In many cases, initial fuzzy sets, each of which has a linguistic meaning, are predefined by domain experts and are thus required to be maintained in order to ensure interpretability of any subsequent inference results. However, learning fuzzy rules using fixed fuzzy quantity space without any quantification will restrict the accuracy of the resulting rules. Fortunately, adjusting the weight of a fuzzy if-then rule may help improve classification accuracy without degrading the interpretability. There have been different proposals for fuzzy rule weight tuning through the use of various heuristics with limited success. This paper proposes an alternative approach using Particle Swarm Optimisation in the search of a set of optimal rule weights, which can entail high classification accuracy. The proposed method is initially tested on the iris data set with regard to different predefined fuzzy partitions of linguistic variables to assess its performance. Experimental results demonstrate that the proposed approach is not sensitive to the predefined fuzzy partitions, and can boost classification performance especially when a coarse fuzzy partition is given.
LanguageEnglish
Title of host publication2014 14th UK Workshop on Computational Intelligence (UKCI)
PublisherIEEE
ISBN (Electronic)9781479955381
DOIs
Publication statusPublished - 20 Oct 2014
Externally publishedYes
Event14th UK Workshop on Computational Intelligence - University of Bradford, Bradford, United Kingdom
Duration: 8 Sep 201410 Sep 2014
Conference number: 14
http://www.computing.brad.ac.uk/ukci2014/ (Link to Conference Website )

Workshop

Workshop14th UK Workshop on Computational Intelligence
Abbreviated titleUKCI 2014
CountryUnited Kingdom
CityBradford
Period8/09/1410/09/14
Internet address

Fingerprint

Fuzzy rules
Particle swarm optimization (PSO)
Fuzzy sets
Linguistics
Tuning

Cite this

Chen, T., Shen, Q., Su, P., & Shang, C. (2014). Refinement of fuzzy rule weights with particle swarm optimisation. In 2014 14th UK Workshop on Computational Intelligence (UKCI) IEEE. https://doi.org/10.1109/UKCI.2014.6930170
Chen, Tianhua ; Shen, Qiang ; Su, Pan ; Shang, Changjing. / Refinement of fuzzy rule weights with particle swarm optimisation. 2014 14th UK Workshop on Computational Intelligence (UKCI). IEEE, 2014.
@inproceedings{68ed8a11aeb64c9b8fad20955fd0cf6a,
title = "Refinement of fuzzy rule weights with particle swarm optimisation",
abstract = "The most challenging problem in the design of fuzzy rule-based classification systems is the construction of a fuzzy rule base for the target problem. Much research has focused on generating and adjusting antecedent fuzzy sets. In many cases, initial fuzzy sets, each of which has a linguistic meaning, are predefined by domain experts and are thus required to be maintained in order to ensure interpretability of any subsequent inference results. However, learning fuzzy rules using fixed fuzzy quantity space without any quantification will restrict the accuracy of the resulting rules. Fortunately, adjusting the weight of a fuzzy if-then rule may help improve classification accuracy without degrading the interpretability. There have been different proposals for fuzzy rule weight tuning through the use of various heuristics with limited success. This paper proposes an alternative approach using Particle Swarm Optimisation in the search of a set of optimal rule weights, which can entail high classification accuracy. The proposed method is initially tested on the iris data set with regard to different predefined fuzzy partitions of linguistic variables to assess its performance. Experimental results demonstrate that the proposed approach is not sensitive to the predefined fuzzy partitions, and can boost classification performance especially when a coarse fuzzy partition is given.",
keywords = "accuracy, fuzzy sets, iris, optimization, pattern matching, pragmatics, training",
author = "Tianhua Chen and Qiang Shen and Pan Su and Changjing Shang",
year = "2014",
month = "10",
day = "20",
doi = "10.1109/UKCI.2014.6930170",
language = "English",
booktitle = "2014 14th UK Workshop on Computational Intelligence (UKCI)",
publisher = "IEEE",

}

Chen, T, Shen, Q, Su, P & Shang, C 2014, Refinement of fuzzy rule weights with particle swarm optimisation. in 2014 14th UK Workshop on Computational Intelligence (UKCI). IEEE, 14th UK Workshop on Computational Intelligence, Bradford, United Kingdom, 8/09/14. https://doi.org/10.1109/UKCI.2014.6930170

Refinement of fuzzy rule weights with particle swarm optimisation. / Chen, Tianhua; Shen, Qiang; Su, Pan ; Shang, Changjing.

2014 14th UK Workshop on Computational Intelligence (UKCI). IEEE, 2014.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Refinement of fuzzy rule weights with particle swarm optimisation

AU - Chen, Tianhua

AU - Shen, Qiang

AU - Su, Pan

AU - Shang, Changjing

PY - 2014/10/20

Y1 - 2014/10/20

N2 - The most challenging problem in the design of fuzzy rule-based classification systems is the construction of a fuzzy rule base for the target problem. Much research has focused on generating and adjusting antecedent fuzzy sets. In many cases, initial fuzzy sets, each of which has a linguistic meaning, are predefined by domain experts and are thus required to be maintained in order to ensure interpretability of any subsequent inference results. However, learning fuzzy rules using fixed fuzzy quantity space without any quantification will restrict the accuracy of the resulting rules. Fortunately, adjusting the weight of a fuzzy if-then rule may help improve classification accuracy without degrading the interpretability. There have been different proposals for fuzzy rule weight tuning through the use of various heuristics with limited success. This paper proposes an alternative approach using Particle Swarm Optimisation in the search of a set of optimal rule weights, which can entail high classification accuracy. The proposed method is initially tested on the iris data set with regard to different predefined fuzzy partitions of linguistic variables to assess its performance. Experimental results demonstrate that the proposed approach is not sensitive to the predefined fuzzy partitions, and can boost classification performance especially when a coarse fuzzy partition is given.

AB - The most challenging problem in the design of fuzzy rule-based classification systems is the construction of a fuzzy rule base for the target problem. Much research has focused on generating and adjusting antecedent fuzzy sets. In many cases, initial fuzzy sets, each of which has a linguistic meaning, are predefined by domain experts and are thus required to be maintained in order to ensure interpretability of any subsequent inference results. However, learning fuzzy rules using fixed fuzzy quantity space without any quantification will restrict the accuracy of the resulting rules. Fortunately, adjusting the weight of a fuzzy if-then rule may help improve classification accuracy without degrading the interpretability. There have been different proposals for fuzzy rule weight tuning through the use of various heuristics with limited success. This paper proposes an alternative approach using Particle Swarm Optimisation in the search of a set of optimal rule weights, which can entail high classification accuracy. The proposed method is initially tested on the iris data set with regard to different predefined fuzzy partitions of linguistic variables to assess its performance. Experimental results demonstrate that the proposed approach is not sensitive to the predefined fuzzy partitions, and can boost classification performance especially when a coarse fuzzy partition is given.

KW - accuracy

KW - fuzzy sets

KW - iris

KW - optimization

KW - pattern matching

KW - pragmatics

KW - training

U2 - 10.1109/UKCI.2014.6930170

DO - 10.1109/UKCI.2014.6930170

M3 - Conference contribution

BT - 2014 14th UK Workshop on Computational Intelligence (UKCI)

PB - IEEE

ER -

Chen T, Shen Q, Su P, Shang C. Refinement of fuzzy rule weights with particle swarm optimisation. In 2014 14th UK Workshop on Computational Intelligence (UKCI). IEEE. 2014 https://doi.org/10.1109/UKCI.2014.6930170