Felix Parker and Darius Irani
AlphaTSP: Learning a TSP Heuristic Using the AlphaZero Methodology
Run with:
python3 main.py --experiment <experiment_name>
Where experiment name is one of:
- nearestneighbor
- mcts
- exact
- gurobi
- insertion
- policy
- parallel
- selfplay
selfplay is the primary experiment which implements the AlphaTSP method
Python packages required to run this project are listed in requirements.txt