Book Image

Deep Reinforcement Learning with Python - Second Edition

By : Sudharsan Ravichandiran
Book Image

Deep Reinforcement Learning with Python - Second Edition

By: Sudharsan Ravichandiran

Overview of this book

With significant enhancements in the quality and quantity of algorithms in recent years, this second edition of Hands-On Reinforcement Learning with Python has been revamped into an example-rich guide to learning state-of-the-art reinforcement learning (RL) and deep RL algorithms with TensorFlow 2 and the OpenAI Gym toolkit. In addition to exploring RL basics and foundational concepts such as Bellman equation, Markov decision processes, and dynamic programming algorithms, this second edition dives deep into the full spectrum of value-based, policy-based, and actor-critic RL methods. It explores state-of-the-art algorithms such as DQN, TRPO, PPO and ACKTR, DDPG, TD3, and SAC in depth, demystifying the underlying math and demonstrating implementations through simple code examples. The book has several new chapters dedicated to new RL techniques, including distributional RL, imitation learning, inverse RL, and meta RL. You will learn to leverage stable baselines, an improvement of OpenAI’s baseline library, to effortlessly implement popular RL algorithms. The book concludes with an overview of promising approaches such as meta-learning and imagination augmented agents in research. By the end, you will become skilled in effectively employing RL and deep RL in your real-world projects.
Table of Contents (22 chapters)
Other Books You May Enjoy

Chapter 10 – Policy Gradient Method

  1. In the value-based method, we extract the optimal policy from the optimal Q function (Q values).
  2. It is difficult to compute optimal policy using the value-based method when our action space is continuous. So, we use the policy-based method. In the policy-based method, we compute the optimal policy without the Q function.
  3. In the policy gradient method, we select actions based on the action probability distribution given by the network and if we win the episode, that is, if we get a high return, then we assign high probabilities to all the actions of the episode, else we assign low probabilities to all the actions of the episode.
  4. The policy gradient is computed as .
  5. Reward-to-go is basically the return of the trajectory starting from the state st. It is computed as .
  6. The policy gradient with the baseline function is a policy gradient method that uses the baseline function to reduce the variance...