Simple reinforcement Learning project without deep learning framework. This is consist of pure C# without any GUI. If you want more detail, take look at my blog post

Menu

1. Evaluate: Evaluating value table by base on reward
2. Improve: Improving policy table by base on value table
3. Move: Making agent move by base on policy table
4. Reset: Currently not working
5. Exit: Exit program

Contents

• Bellman Equation
• Dynamic Programming
• AI
• FrozenLake Game

Environment

• .Net Framework 'v4.6.1'
• Window 10