The Community for Technology Leaders
RSS Icon
Subscribe
Los Angeles, CA
March 31, 2009 to April 2, 2009
ISBN: 978-0-7695-3507-4
pp: 250-255
ABSTRACT
Fuzzy relational classifier (FRC) is the recently proposed two-step nonlinear classifiers, which effectively integrates the formed clusters and the given classes. However, FRC can not copy with the influence of  those irrelevant or redundant features. To effectively filter out those irrelevant features and  preserve the internal structure hidden in the given data, in this paper, a simultaneous clustering and feature selection framework called FCM_FS is introduced, which incorporates margin based feature selection  criterion into the unsupervised fuzzy c-means(FCM) clustering. Based on  FCM_FS  and FRC framework, we introduce an enhanced FRC (EFRC). The experimental results on 8 real-life benchmark datasets show that: EFRC can consistently outperform FRC in classification performance.
INDEX TERMS
FCM, Classification, Feature Selection, Enhanced Fuzzy Relational Classifier (EFRC)
CITATION
Ming Yang, Jing Song, Gen-lin Ji, "FCM_FS: A Simultaneous Clustering and Feature Selection Model for Classification", CSIE, 2009, 2009 WRI World Congress on Computer Science and Information Engineering, CSIE, 2009 WRI World Congress on Computer Science and Information Engineering, CSIE 2009, pp. 250-255, doi:10.1109/CSIE.2009.347
4 ms
(Ver 2.0)

Marketing Automation Platform Marketing Automation Tool