An in-depth analysis of fuzzy c-means clustering for cellular manufacturing

Jie Li, Chao Hsien Chu, Yunfeng Wang

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

2 Scopus citations

Abstract

Fuzzy c-means (FCM), a well-known clustering algorithm, has been successfully adapted to solve a variety of applications including cellular manufacturing. This paper provides an in-depth analysis on the deficiencies of applying FCM to solve the cell formation (CF) problem in cellular manufacturing and proposes ways of enhancing its performance. A large-scale experiment is conducted to evaluate the effects of different enhancements over FCM. Our study shows that, for CF problem, (1) the proposed distance function has the largest impact on solution quality, followed by the subtractive initialization, (2) the effects of center function and solution selection are not as significant as the formers, and (3) combining the proposed distance function and subtractive initialization with FCM produces the most synergic effects in improving solution quality, while only adding a tolerable amount of computation time.

Original languageEnglish (US)
Title of host publicationProceedings - 5th International Conference on Fuzzy Systems and Knowledge Discovery, FSKD 2008
Pages42-46
Number of pages5
DOIs
StatePublished - 2008
Event5th International Conference on Fuzzy Systems and Knowledge Discovery, FSKD 2008 - Jinan, Shandong, China
Duration: Oct 18 2008Oct 20 2008

Publication series

NameProceedings - 5th International Conference on Fuzzy Systems and Knowledge Discovery, FSKD 2008
Volume1

Other

Other5th International Conference on Fuzzy Systems and Knowledge Discovery, FSKD 2008
Country/TerritoryChina
CityJinan, Shandong
Period10/18/0810/20/08

All Science Journal Classification (ASJC) codes

  • Artificial Intelligence
  • Control and Systems Engineering

Fingerprint

Dive into the research topics of 'An in-depth analysis of fuzzy c-means clustering for cellular manufacturing'. Together they form a unique fingerprint.

Cite this