Reinforcement learning for ART-based fuzzy adaptive learning control networks

摘要

This paper proposes a reinforcement fuzzy adaptive learning control network (RFALCON) for solving various reinforcement learning problems. The proposed RFALCON is constructed by integrating two fuzzy adaptive learning control networks (FALCON), each of which is a connectionist model with a feedforward multilayered network developed for the realization of a fuzzy logic controller. An online structure/parameter learning algorithm, called RFALCON-ART, is proposed for constructing the RFALCON dynamically. The proposed RFALCON also preserves the advantages of the original FALCON, such as the ability to do online partition the input/output spaces, tune membership functions, and find proper fuzzy logic rules. In its initial form, there is no membership function, fuzzy partition, and fuzzy logic rule. They are created and begin to grow as the first reinforcement signal arrives. The users thus need not give it any a priori knowledge or even any initial information on these.