Continuous Deep Q-Learning with Model-based Acceleration

   Abstract