Bachelor's Project
For my bachelors project, I used reinforcement learning to train an agent in controlling forest fires. In order to do so, me and three other students from the University of Groningen build a forest fire simulation and implemented several reinforcement learning techniques. My main contributions were the agent implementations and the connectionist monte- Carlo algorithm inplementation. The final grade awarded to this project was a 7.5.
The source code can be found here. Due to the lack of running instructions and proper documentation, the thesis might be more informative