Fuzzy Mechanisms For Causal Relations

Abstract

Classical fuzzy logic does not allow the implementation of causal relations as defined in causal maps. Until now, Fuzzy Causal Maps(FCM) have been implemented by the use of mechanisms closer to neural networks that can not be mixed with classic fuzzy rule based systems. Rule based fuzzy cognitive maps importance resides in the fact that they could allow a more complete representation of cognition since relations other than causality would be possible. This paper presents a method to implement Fuzzy Causal Relations that can be used in Rule Based Fuzzy Cognitive Maps (RBFCM). This method introduces a new fuzzy operation that simulates the “accumulative” property associated with causal relations – the Fuzzy Carry Accumulation (FCA). The FCA allows a great flexibility in the addition and removal of concepts and links among concepts while keeping compatibility with classic fuzzy operations.