Daisuke Kitakoshi, Hiroyuki Shioya and Tsutomu Da-te
A Reinforcement Learning Method Supported by a Bayesian Network