DeepRead
Cover of Grokking Deep Reinforcement Learning

Grokking Deep Reinforcement Learning

Miguel Morales · 2020

Machine learning, statistics and probability › Deep learning and AI

#8,691 most recommended

About the book

An introduction to deep reinforcement learning, covering how an agent learns to make sequential decisions by interacting with an environment to maximize reward. It starts from foundations such as multi-armed bandits and Markov decision processes, then moves to value-based and policy-based methods that use deep neural networks, with Python and PyTorch code examples. The book's position is that deep reinforcement learning is best learned by building intuition step by step, with each algorithm presented as a small, visually illustrated extension of the one before it. It holds that the core ideas (value functions, the Bellman relationship, exploration versus exploitation, and policy gradients) recur across methods, so understanding them makes later algorithms such as DQN and actor-critic methods easier to follow. It supports these claims with runnable implementations rather than relying only on mathematical derivations.

Written by Claude from what it knows of this book.

What Hacker News says 33% recommend · 0% critical

02420222023202420252024 peak · 3
This bookAverage book in Deep learning and AI

In the commenters’ own words

What Open Library readers say 5.00 from 1 rating

  1. 51
  2. 40
  3. 30
  4. 20
  5. 10
  • 4 want to read
  • 1 reading
  • 1 have read

Only 1 rating: treat the average as a hint.

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
2020
ISBN
9781638356660
Editions known
4 at Open Library
Open Library
OL25779058W

Explore in the mosaic

Read next: its neighbours on the mosaic