Exploiting Multiple Abstractions in Episodic RL via Reward Shaping

Roberto Cipollone‚ Giuseppe De Giacomo‚ Marco Favorito‚ Luca Iocchi and Fabio Patrizi