Designing an efficient fuzzy classifier using an intelligent genetic algorithm

Shinn-Ying Ho*, Tai Kang Chen, Shinn Jang Ho

*Corresponding author for this work

Research output: Contribution to journalConference article

7 Scopus citations


In this paper, we propose a method for designing an efficient fuzzy classifier that consists of a small number of fuzzy rules with only a few antecedent fuzzy sets using a novel intelligent genetic algorithm (IGA). It is known that the number of fuzzy rules will be exploded as the number of features increases. So the fuzzy classifier with many input variables has extremely large number of fuzzy rules for high-dimensional pattern classification problems. To cope with this large rule base problem, our proposed method has the following three merits: (1) A flexible genetic parameterized fuzzy region is proposed to efficiently partition the feature space. (2) The parametric genes for representing the membership functions and fuzzy rules, and the control genes used for useful pattern feature selection and dummy fuzzy rule deletion are incorporated into a single chromosome. This means that the participated features, the membership function of each antecedent fuzzy set, and the fuzzy rules are simultaneously determined. (3) The efficient fuzzy classifier design is formulated as a large parameter optimization problem (LPOP). We solve LPOPs using a novel IGA which is superior to the conventional genetic algorithms in solving LPOPs. The high performance of the proposed method is illustrated by computer simulations on the iris and wine classification problems and the simulation results are superior to those of the existing methods.

Original languageEnglish
Pages (from-to)293-298
Number of pages6
JournalProceedings - IEEE Computer Society's International Computer Software and Applications Conference
StatePublished - 1 Dec 2000
Event2000 IEEE 24th Annual International Computer Software and Applications Conference (COMPSAC 2000) - Taipei, Taiwan
Duration: 25 Oct 200027 Oct 2000

Fingerprint Dive into the research topics of 'Designing an efficient fuzzy classifier using an intelligent genetic algorithm'. Together they form a unique fingerprint.

  • Cite this