Reinforcement Learning with Python
Reinforcement Learning with Python


This book is no longer available for sale.

Reinforcement Learning with Python

This book is 100% complete

Completed on 2017-12-20

About the Book

Reinforcement Learning is regarded by many as the next big thing in data science. Beyond the hype, there is an interesting, multidisciplinary and very rich research area, with many proven successful applications, and many more promising. In this book I will introduce the main tools, ideas and history of the field, with a mixture of theory and practice (using examples from OpenAI Gym). Constantly in progres!!

About the Author

Pablo Maldonado
Pablo Maldonado

Pablo is an applied mathematician and data scientist with a taste for software development since his days programming BASIC in a Tandy 1000.

He spends a great deal of his time building machine learning products and helping others succeed in their own data-driven projects, whether companies or students.

You can follow him in where he blogs about data science and math in the wild.

Table of Contents

  • Introduction: Reinforcement Learning
  • Some RL success stories
  • Welcome to Reinforcement Learning
    • An example: OpenAI Gym
    • Using a different policy
    • Your turn:
  • Markov Decision Processes
    • Markov chains
    • Markov Reward Process
    • Markov Decision Processes
    • Solving MDPs: Value and Policy Iteration.
  • Chapter 3: Monte Carlo Methods
    • Monte Carlo Learning
    • Off-policy MC control
  • Chapter 4: Temporal Difference Learning
    • Code sample: SARSA
  • Chapter 5: Function approximation.
    • Gradient descent
    • Feature vectors
    • Function backups
    • Code sample: SARSA with linear function approximation
  • Chapter 6: Experience Replay.
    • Improvements since the original DQN
    • #
    • Code sample: Q-Learning with experience replay (Linear Approximator)
    • Code sample: Q-Learning with experience replay (Neural Network Approximator)
  • Chapter 7: Policy Gradients & Policy Optimisation
    • Derivative Free Methods
    • Code sample: NES for FrozenLake
    • Code sample: Cross entropy method for CartPole
  • About the author

The Leanpub 45-day 100% Happiness Guarantee

Within 45 days of purchase you can get a 100% refund on any Leanpub purchase, in two clicks.

See full terms...

Write and Publish on Leanpub

Authors, publishers and universities use Leanpub to publish amazing in-progress and completed books and courses, just like this one. You can use Leanpub to write, publish and sell your book or course as well! Leanpub is a powerful platform for serious authors, combining a simple, elegant writing and publishing workflow with a store focused on selling in-progress ebooks. Leanpub is a magical typewriter for authors: just write in plain text, and to publish your ebook, just click a button. It really is that easy.

Learn more about writing on Leanpub