David-Silver-Reinforcement-learning

Why we included this project

Watching David Silver's reinforcement learning lectures is one thing; getting the algorithms to actually run is another. This repository pairs notes that follow the full course, from Markov decision processes and dynamic programming through to policy gradients and exploration versus exploitation, with working Keras implementations that run on OpenAI Gym. Each algorithm comes with the notes that explain it, so you can read the idea and then run it against a concrete environment. Students and self-taught engineers who learn better by running code than by watching slides will find value iteration, Q-learning, and policy gradient methods implemented rather than just described. When a concept fades, the repository also works as a quick reference to come back to.