On modeling similarity and three-way decision under incomplete information in rough set theory

摘要: Abstract Although incomplete information is a well studied topic in rough set theory, there still does not exist general agreement on the semantics of various types information. This has led to some confusions and many definitions similarity or tolerance relations objects, without sound semantical justification. The main objective this paper address issues related We present four-step model Pawlak analysis, order gain insights how an indiscernibility relation (i.e., equivalence relation) defined used under complete results enable us propose conceptual framework for studying objects based classification four “do-not-care value”, “partially-known “class-specific “non-applicable value”) two groups methods relation-based granule-based methods) modeling similarity. examine existing studies their relationships. In spite differences, all can be uniformly represented set-valued table. are therefore able have common possible-world semantics. Finally, demonstrate value proposed framework, we three-way decisions