Dynamic Fuzzy Logic

This paper presents such a way of fuzzy inferencing that a resulting judgment changes in time. The research was motivated by the empirical evidence that people’s feelings about a perceived person or a social situation may oscillate between a highly positive value and a highly negative value even in absence of new data. The proposed method of judgment generation employs the classic fuzzy-set-theoretic concept of membership. A set of membership values is used as the control parameter a in a dynamic model Xt+1 = F(a, Xt ), where Xt is a set of state variables. One of the state variables can be interpreted as resulting judgment based on 'feelings' given as a dynamic membership. The function F can be implemented as a feed-forward neural network.