A neural architecture to address Reinforcement Learning problems | IEEE Conference Publication | IEEE Xplore