deep Reinforcement Learning Hands-On: Apply modern RL methods, with Q-networks, value iteration, policy gradients, TRPO, AlphaGo Zero and more