paper: https://arxiv.org/abs/2510.20235

1. conda env create -f environment.yaml
2. conda activate mo_traffic
3. sudo add-apt-repository ppa:sumo/stable
4. sudo apt-get update
5. sudo apt-get install sumo sumo-tools sumo-doc
6. echo 'export SUMO_HOME="/usr/share/sumo"' >> ~/.bashrc
7. source ~/.bashrc

For more information about SUMO, please refer to: https://github.com/LucasAlegre/sumo-rl

Example commands for running the algorithms are as follows:

• ARAM

  python maxmin_algorithms.py -ename base4 -rd 4 -tt 100000 -nepoch 8 -nsteps 128 -bsppo 32 -mlr 0.001 -lam 0.1 -beta 0.67 -se 0

• ERAM

  python maxmin_algorithms.py -useci -ename base4 -rd 4 -tt 100000 -nepoch 8 -nsteps 128 -bsppo 32 -mlr 0.001 -lam 0.2 -beta 0.67 -se 0

If you find this work useful, please cite:

@inproceedings{byeonmulti,
  title={Multi-Objective Reinforcement Learning with Max-Min Criterion: A Game-Theoretic Approach},
  author={Byeon, Woohyeon and Park, Giseung and Chae, Jongseong and Leshem, Amir and Sung, Youngchul},
  booktitle={The Thirty-ninth Annual Conference on Neural Information Processing Systems}
}

woohyeon.byeon@kaist.ac.kr