Home

• Theoretical overview
• Implementation analysis
• Practical example visual results

Purpose

Application of this Q-Learning module to a dynamic fire approach.

Reasons

• This project offers the Q-Learning algorithm implemented without using any AI library, in order to make it easy to experiment and personalize it.

• The hashed instance of the State class and the use of lists, sets and tuples instead of Numpy arrays provides a much faster learning convergence process.

• This project also offers unique visualization approaches of the learning convergence process.