This repo contains my experiments with RL algorithms
- Solve maze with DQN, represent state as (x, y), use hidden layers
- Dynamic maze with an agent
- Solve tic-tac-toe with Q-tables
- Solve tic-tac-toe with DQN
- Solve continuous time physics problem
- Solve lunar-lender
- Solve inverted pendulum
- train NN to predict simulation (use it as a model later)
- RL tutorial
- https://habr.com/ru/post/483078/
- https://habr.com/ru/post/470179/
- https://habr.com/ru/post/345656/
- http://ai.berkeley.edu/reinforcement.html
- https://habr.com/ru/post/442522/
- Update conda environment:
conda env update --file environment.yml --prune