Fuzzy Relations and Fuzzy Logic Inference
David B. Fogel, Derong Liu, James M. Keller · 2016
This chapter describes the background necessary to understand and construct fuzzy logic inference systems for decision-making problems and control applications. Fuzzy logic begins with the concept of a linguistic variable. Once we have this fundamental concept of a linguistic variable, we can build the machinery necessary for fuzzy logic inference. The chapter provides mechanisms for making deductions that are all based on the concept of fuzzy relations. There are many examples of direct applications of fuzzy relations, and the chapter concentrate on the main use, that of providing an engine for logical inference in a fuzzy rule-based system. IF-THEN rules form the basis of a fuzzy logic inference system. Systems of fuzzy rules can be built or learned to perform control functions, but also to work as a pattern classifier. These systems in many cases behave like statistical classifiers, but can also encode human knowledge directly into the structure in a linguistically pleasing manner.