Key points are not available for this paper at this time.
Reinforcement learning aims to determine an optimal control policy from interaction with a system or from observations gathered from a system. In batch mode, it can be achieved by approximating the...
ErnstDamien et al. (2005) studied this question.