DeepRead
Cover of Deep Reinforcement Learning Hands-On

Deep Reinforcement Learning Hands-On

Maxim Lapan · 2018

Machine learning, statistics and probability › Deep learning and AI

#11,358 most recommended

About the book

A practical introduction to deep reinforcement learning, built around PyTorch and the OpenAI Gym toolkit. It covers the core methods from simple tabular and cross-entropy approaches through deep Q-learning and policy-gradient and actor-critic methods, applied to classic control tasks and Atari games. It also includes applications beyond games, such as web navigation and discrete optimization problems. The book holds that deep reinforcement learning is best learned by implementing the algorithms step by step, starting from small, easily checked environments before moving to harder ones. It presents value-based, policy-based, and actor-critic methods as related ways of estimating and using expected returns, rather than unrelated techniques. It also stresses that practical details, such as hyperparameters, stabilization tricks, and debugging training runs, often decide whether a method works in practice.

Written by Claude from what it knows of this book.

What Hacker News says 33% recommend · 0% critical

01220182020202220232018 peak · 1
This bookAverage book in Deep learning and AI

In the commenters’ own words

What Open Library readers say

No Open Library reader has rated it yet.

  • 8 want to read
  • 0 reading
  • 0 have read

Open Library’s page for the book

What critics say, via Wikipedia

No Wikipedia article about this book yet, so no critics to quote.

Reddit communities — coming later

Details

First published
2018
ISBN
9781788834247
Editions known
8 at Open Library
Open Library
OL25246516W

Explore in the mosaic

Read next: its neighbours on the mosaic