Deep Reinforcement Learning with Explicit Context Representation