Provided by: International Journal of Computer Applications
Date Added: Nov 2012
This paper presents Learning Fuzzy Cognitive Maps (LFCM) as a new paradigm, or approach, for modeling Complex Adaptive Systems (CAS). This technique is the fusion of the advances of the fuzzy logic, formal neural network, and reinforcement learning where they are suitable for modeling systems in artificial life domain of CAS. The FCM structure is similar to a recurrent artificial neural network. The Reinforcement Learning (RL) gives the explicative frame of entities like environment changing adaptation. A mathematical adaptation of the Q-learning algorithm is discussed and the authors present in this paper an inspired pseudo-hybridization algorithm Q-learning, mainly used in non-linear dynamic systems RL, and the Hebb law for the inference calculus introduced by the cognitive maps. The prey and predator simulation model is shown.