FS-FOIL: An inductive learning method for extracting interpretable fuzzy descriptions

Autoren Mario Drobics
Ulrich Bodenhofer
Erich Peter Klement
TitelFS-FOIL: An inductive learning method for extracting interpretable fuzzy descriptions
TypArtikel
JournalInternational Journal of Approximate Reasoning
Nummer2-3
Band32
Noteinvited
AbteilungKVS
ISSN0888-613X
MonatFebruary
Jahr2003
Seiten131-152
SCCH ID#201
Abstract

This paper is concerned with FS-FOIL - an extension of Quinlan's First-Order Inductive Learning Method (FOIL). In contrast to the classical FOIL algorithm, FS-FOIL uses fuzzy predicates and, thereby, allows to deal not only with categorical variables, but also with numerical ones, without the need to draw sharp boundaries. This method is described in full detail along with discussions how it can be applied in different traditional application scenarios - classification, fuzzy modeling, and clustering. We provide examples of all three types of applications in order to illustrate the efficiency, robustness, and wide applicability of the FS-FOIL method.