Autonomous Synthesis of Fuzzy Cognitive Maps from Observational Data: Preliminaries

We review current research in fuzzy cognitive map (FCM) parameter estimation techniques and present results for a method based on constrained optimization as motivation for a discussion of the core issue of causal discovery in intelligent systems. While FCM parameter estimation techniques show utility for automatically determining causal strength from state observations, their success relies heavily on background knowledge about the causal structure of the system, acquired from interviews with domain experts. In this paper we identify causal discovery as a fundamental task for automated FCM synthesis and present a survey of related research in automatic causal discovery that provides a basis for future work in the area of autonomous FCM synthesis