Rough Sets and Knowledge Discovery: An Overview

Abstract

The primary methodological framework to study classification problems with imprecise or incomplete information in this book is the theory of rough sets. The theory was originally introduced by Pawlak[1]. The uniqueness as well as the complementary character of rough set theory to other approaches for dealing with imprecise, noisy, or incomplete information such as fuzzy set theory[4], or theory of evidence[5] was recognized by mathematicians and researchers working on mathematical foundations of Computer Science. Currently, there are over 800 publications in this area, including two books and an annual workshop. The rough sets model is used as a departure point to study formal reasoning with uncertain information[6–8], machine learning, knowledge discovery[9–13, 20], and representation and reasoning about imprecise knowledge[6]. The theory of rough sets has been applied in numerous domains such as, for example, analysis of clinical data and medical diagnosis[14], information retrieval[15], control algorithm acquisition and process control[16], analysis of complex chemical compounds[17], structural engineering[18], market analysis[12], and others[9].