An Acquiring Method of Macro-Actions in Reinforcement Learning | IEEE Conference Publication | IEEE Xplore