Purpose: evaluate publically available, open-source environments for similarities with Naval problem areas, and implement a variety of Reinforcement Learning algorthims to develop optimal strategies. Enviroments will be adjusted to alter reward stuctures, environmental cues and complexity to better represent real-world situations
- Python Learning Environment: WaterWorld
- Algorithm(s): Deep Q-Learning
- Python 3.7.3
- pandas 0.25.3
- numpy 1.17.4
- ple 0.0.1
- matplotlib 3.1.1
- keras 2.2.4
- keras-rl 0.4.2
- tensorflow 1.15.0