AN APPROACH TO CAUSAL-MODELING IN FUZZY ENVIRONMENT AND ITS APPLICATION

This paper suggests an approach to building a causal model in fuzzy environment. The fuzzy vector autoregressive (FVAR) model is defined and a causality definition in the fuzzy environment is provided. Three types of causality, namely, nonfuzzy non-causality, nonfuzzy causality, and fuzzy causality are defined based on the fuzzy causality relation introduced in this paper. Necessary and sufficient conditions for the three types of causality are also provided when fuzzy parameters have triangular membership functions. A procedure to implement the fuzzy causal modeling is also suggested. We applied the procedure to real data and compared it with other conventional empirical studies. © 1990.