Fuzzy cognitive map and its causal inferences

Fuzzy cognitive maps (FCM) is a powerful framework for representing structured human knowledge and causal inference. This paper presents a new and effective approach to analyzing causal inference mechanism of FCM. We focus on binary concept states proposed originally by Kosko (1986). Given initial conditions, FCM is able to reach only certain states in its state space. The problem of finding whether a state is reachable in the FCM is NP hard. In order to effectively carry out the design of fuzzy cognitive maps we propose to divide an FCM into basic FCM modules. This paper also presents a recursive formula for the calculation of FCM inference patterns in terms of key vertices.