Rough sets and decision rules in fuzzy set-valued information systems

Abstract

Set-valued Information Systems(SVISs) are generalized forms of Crisp Information Systems(CISs) and common in practice. This paper defines a fuzzy inclusion relation in Fuzzy Set-valued Information Systems(FSVISs). By means of two parameters of inclusion degree λ1 and λ 2, we define the rough sets in FSVISs, which are used to approximate fuzzy concepts in FSVISs. Furthermore, in terms of the maximum elements in the lattice derived from the universe according to decision attributes, we present the definitions and measuring methods of decision rules in FSVISs. Some examples have been given for illustration. © Springer-Verlag Berlin Heidelberg 2005.