Comparing Fuzzy-Rough and Fuzzy Entropy-assisted Fuzzy-Rough Feature Selection

Research output: Contribution to conferencePaper

20 Downloads (Pure)

Abstract

Feature Selection (FS) methods based on fuzzy-rough set theory (FRFS) have employed the dependency function to guide the FS process with much success. More recently a method has been developed which uses fuzzy-entropy \cite{mjs06} to perform this task. Such use of fuzzy-entropy as an evaluation measure in fuzzy-rough feature selection can result in smaller subset sizes than those obtained through FRFS alone. However, it has also been observed that the fuzzy-entropy based FS technique (which does not select subsets based on dependency), also demonstrates remarkably similar dependency values to those of the fuzzy-rough method. This paper investigates the apparent similarity of the dependency values and attempts to discover if any correlation exists. Results are obtained using both fuzzy-rough FS (which is guided solely by the dependency value) and the fuzzy entropy-assisted fuzzy-rough FS technique.
Original languageEnglish
Publication statusPublished - 04 Sept 2006
EventProceedings of the 6th Annual Workshop on Computational Intelligence (UKCI'06) - Reading, United Kingdom of Great Britain and Northern Ireland
Duration: 28 May 200631 May 2006

Conference

ConferenceProceedings of the 6th Annual Workshop on Computational Intelligence (UKCI'06)
Country/TerritoryUnited Kingdom of Great Britain and Northern Ireland
Period28 May 200631 May 2006

Fingerprint

Dive into the research topics of 'Comparing Fuzzy-Rough and Fuzzy Entropy-assisted Fuzzy-Rough Feature Selection'. Together they form a unique fingerprint.

Cite this