The aim of this paper is to develop an AI agent with self-learning capabilities that is able to play classical Atari console games without human intervention and achieve next to human level performance. In order to achieve ...