IJCNN
2008
IEEE
14 years 6 months ago
2008
IEEE
— In this paper we address the reliability of policies derived by Reinforcement Learning on a limited amount of observations. This can be done in a principled manner by taking in...