We are developing a new reinforcement learning (RL) technique that has the mechanism for segmenting continuous state space and continuous action space autonomously. This is a kind ofinstance-basedReinforcement Learning (BRL). We have presented elsewhere that the BRL is an effective technique for Mu]ti-Robot Systems (MRS) through several experiments. However, as most learning systems do, BRL owns the problem ofover-fitting. This paper shows an extension of BRL in order to overcome this problem. The idea is that rules which are fbund to be effectiye in the process oflearning are protected from the rule deletion. This sirnple strategy makes
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.