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)
18
Other Books You May Enjoy
19
Index

Policy gradient intuition

Policy gradient is one of the most popular algorithms in deep reinforcement learning. As we have learned, policy gradient is a policy-based method by which we can find the optimal policy without computing the Q function. It finds the optimal policy by directly parameterizing the policy using some parameter .

The policy gradient method uses a stochastic policy. We have learned that with a stochastic policy, we select an action based on the probability distribution over the action space. Say we have a stochastic policy , then it gives the probability of taking an action a given the state s. It can be denoted by . In the policy gradient method, we use a parameterized policy, so we can denote our policy as , where indicates that our policy is parameterized.

Wait! What do we mean when we say a parameterized policy? What is it exactly? Remember with DQN, we learned that we parameterize our Q function to compute the Q value? We can do the same here, except...