Deep Reinforcement Learning Tutorials

We provide here a set of beginner hands-on Python examples in both Classical & Deep Reinforcement Learning. Each example is self-contained with all code stored in one .py file, and usage of it involves only running the file in its local directory. Each example has its own README.md file briefly discussing the theory and usage.

Classical Reinforcement Learning

• Solving the Multi-Armed Bandit Problem (multiarmed_bandit.py)

• Training an Agent to beat GridWorld (gridworld.py)

• Solving the Hunter-Prey problem as a single-agent problem using relative coordinates (hunterprey.py)

Deep Reinforcement Learning

• Training an Agent to beat GridWorld using Deep Q-Networks (gridworld_DQN.py)