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

A2C revisited

We can design our A2C algorithm with many worker agents, just like the A3C algorithm. However, unlike A3C, A2C is a synchronous algorithm, meaning that in A2C, we can have multiple worker agents, each interacting with their own copies of the environment, and all the worker agents perform synchronous updates, unlike A3C, where the worker agents perform asynchronous updates.

That is, in A2C, each worker agent interacts with the environment, computes losses, and calculates gradients. However, it won't send those gradients to the global network independently. Instead, it waits for all other worker agents to finish their work and then updates the weights to the global network in a synchronous fashion. Performing synchronous weight updates reduces the inconsistency introduced by A3C.