
- Home
- Books
- Artificial Intelligence & Machine Learning
- Reinforcement Learning
- Deep Reinforcement Learning Hands-On
Deep Reinforcement Learning Hands-On by Maxim Lapan is a practical guide to training decision-making neural agents with PyTorch and OpenAI Gymnasium. Reinforcement learning enables agents to solve sequential decision problems in games, robotics, and financial trading. Lapan demonstrates how to implement state-of-the-art algorithms from raw tensor calculations to complete trained models.
What you will learn
- Implement core MDP concepts, Bellman equations, and value iteration algorithms
- Build Deep Q-Networks (DQN) with target networks, experience replay, and dueling layers
- Train policy gradient methods including REINFORCE, Advantage Actor-Critic (A2C), and A3C
- Master advanced continuous control algorithms such as PPO, TRPO, and SAC
- Apply reinforcement learning agents to Atari games, robotic simulations, and financial markets
- Scale training throughput using parallel environments and vectorized PyTorch batches
Who this is for
Machine learning engineers, researchers, and data scientists wanting to master practical deep reinforcement learning. Practical proficiency in Python and basic PyTorch knowledge is assumed. Prior reinforcement learning experience is not required.


