Implementing fuzzy constraints in a programming environment

The central focus of this work is the investigation of the use of mathematical programming methods as a tool for implementing the reasoning (inference) in fuzzy logics. Since the general paradigm for reasoning in fuzzy logic is based upon the satisfaction of a conjunction of propositions represented as fuzzy (elastic) constraints, mathematical programming appears to provide a natural framework to implement this reasoning process. A central concern here is the issue of tractability of knowledge representation in the MP domain. In particular, while fuzzy logic provides a large collection of tools for translating natural language propositions into fuzzy constraints, these constraints may be highly nonlinear. The author is interested in providing these fuzzy constraints in a manner amenable to representation and solution by available MP codes. The author calls this the problem of knowledge representation in the MP domain