On Training-Test (Mis)alignment in Unsupervised Combinatorial Optimization

This repository contains the code for the paper "On Training-Test (Mis)alignment in Unsupervised Combinatorial Optimization".

This work is accepted to the second workshop on test-time adaptation at ICML 2025.

The toy_example folder contains the code for the experiments on the toy example.

The facility_location folder contains the code for the experiments on facility location.

Please check the README file in each folder for the guidance for the experiments on each problem.

References

If you find this work useful, please consider citing:

@inproceedings{Bu2025UCOAlign,
  title={On Training-Test (Mis)alignment in Unsupervised Combinatorial Optimization: Observation, Empirical Exploration, and Analysis},
  author={Bu, Fanchen and Shin, Kijung},
  booktitle={ICML Workshop on Test-Time Adaptation},
  year={2025}
}

The paper is available on Arxiv.

You can also check the GitHub Repository for our previous work on unsupervised combinatorial optimization.

@inproceedings{Bu2024UCom2,
  title={Tackling Prevalent Conditions in Unsupervised Combinatorial Optimization: Cardinality, Minimum, Covering, and More},
  author={Bu, Fanchen and Jo, Hyeonsoo and Lee, Soo Yong and Ahn, Sungsoo and Shin, Kijung},
  booktitle={ICML},
  year={2024}
}