Rough Set Based Uncertain Knowledge Expressing and Processing *

Speaker: Guoyin Wang, Institute of Computer Science and Technology, Chongqing University of Posts and Telecommunications, Chongqing, 400065, China P. R.

e-mail: wanggy@ieee.org

ABSTRACT

Uncertainty exists almost everywhere in the whole world. Uncertain knowledge expressing and processing has become one of the most important key problems of artificial intelligence research. There are many kinds of uncertainties in knowledge, such as randomness, fuzziness, vagueness, incompleteness, inconsistency, etc. Randomness and fuzziness are the two most important and fundamental ones. There are many studies about randomness and fuzziness in the past decades. Many theories and models for expressing and processing uncertain knowledge, such as probability & statistics, fuzzy set, rough set, interval analyses, cloud model, grey system, set pair analyses, extenic, etc have been proposed.

 In this talk, some key expanded set theories for expressing and processing uncertain knowledge, such as fuzzy set, rough set, type-II fuzzy set, interval-valued fuzzy set, intuitionistic fuzzy set, and cloud model are discussed. Their key idea and basic notions are introduced. Their difference and relationship are further analyzed.

Rough set theory, which expresses and processes uncertain knowledge with certain methods, is discussed in detail. At first, the growing history of rough set theory is introduced briefly, and the developing trend of rough set theory is analyzed in several views. Then, the expansion of rough set theory to classical set theory is explained. The key set operators of rough set theory, such as intersection, union, difference, and complement, are explained with notions of classical set theory. Rough logic defined on information systems is also analyzed. Several typical application cases of rough set theory in artificial intelligence fields, such as fault diagnosis, intelligent decision, image processing, huge data processing, intelligent control and etc., are discussed to show the power of rough set for dealing with real world problems. These application cases illustrate the importance and advantages of rough set theory for expressing and processing the uncertain problems.

At last, some key topics and problems to be further studied in the future for expressing and processing uncertain knowledge are discussed.

Key words: uncertain knowledge expressing, uncertain knowledge processing, fuzzy set, rough set, cloud model

* This work is supported by National Natural Science Foundation of P. R. China under grant 61073146, Natural Science Foundation Project of CQ CSTC under grant 2008BA2041.