Foundations of Deep Reinforcement Learning: Theory and Practice in Python

Graesser, Laura, Keng, Wah Loon



In just a few years, deep reinforcement learning (DRL) systems such as DeepMinds DQN have yielded remarkable results. This hybrid approach to machine learning shares many similarities with human learning: its unsupervised self-learning, self-discovery of strategies, usage of memory, balance of exploration and exploitation, and its exceptional flexibility. Exciting in its own right, DRL may presage even more remarkable advances in general artificial intelligence.
Deep Reinforcement Learning in Python: A Hands-On Introduction is the fastest and most accessible way to get started with DRL. The authors teach through practical hands-on examples presented with their advanced OpenAI Lab framework. While providing a solid theoretical overview, they emphasize building intuition for the theory, rather than a deep mathematical treatment of results. Coverage includes:

  • Components of an RL system, including environment and agents
  • Value-based algorithms: SARSA, Q-learning and extensions, offline learning
  • Policy-based algorithms: REINFORCE and extensions; comparisons with value-based techniques
  • Combined methods: Actor-Critic and extensions; scalability through async methods
  • Agent evaluation
  • Advanced and experimental techniques, and more


Laura Graesser is a research software engineer working in robotics at Google. She holds a master's degree in computer science from New York University, where she specialized in machine learning.

Wah Loon Keng is an AI engineer at Machine Zone, where he applies deep reinforcement learning to industrial problems. He has a background in both theoretical physics and computer science.