Balancing Interference and Correlation in Spatial Experimental Designs: A Causal Graph Cut Approach

Jin Zhu, Jingyi Li, Hongyi Zhou, Yinan Lin, Zhenhua Lin, Chengchun Shi
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:79918-79937, 2025.

Abstract

This paper focuses on the design of spatial experiments to optimize the amount of information derived from the experimental data and enhance the accuracy of the resulting causal effect estimator. We propose a surrogate function for the mean squared error (MSE) of the estimator, which facilitates the use of classical graph cut algorithms to learn the optimal design. Our proposal offers three key advances: (1) it accommodates moderate to large spatial interference effects; (2) it adapts to different spatial covariance functions; (3) it is computationally efficient. Theoretical results and numerical experiments based on synthetic environments and a dispatch simulator that models a city-scale ridesharing market, further validate the effectiveness of our design. A python implementation of our method is available at https://github.com/Mamba413/CausalGraphCut.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-zhu25l, title = {Balancing Interference and Correlation in Spatial Experimental Designs: A Causal Graph Cut Approach}, author = {Zhu, Jin and Li, Jingyi and Zhou, Hongyi and Lin, Yinan and Lin, Zhenhua and Shi, Chengchun}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {79918--79937}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/zhu25l/zhu25l.pdf}, url = {https://proceedings.mlr.press/v267/zhu25l.html}, abstract = {This paper focuses on the design of spatial experiments to optimize the amount of information derived from the experimental data and enhance the accuracy of the resulting causal effect estimator. We propose a surrogate function for the mean squared error (MSE) of the estimator, which facilitates the use of classical graph cut algorithms to learn the optimal design. Our proposal offers three key advances: (1) it accommodates moderate to large spatial interference effects; (2) it adapts to different spatial covariance functions; (3) it is computationally efficient. Theoretical results and numerical experiments based on synthetic environments and a dispatch simulator that models a city-scale ridesharing market, further validate the effectiveness of our design. A python implementation of our method is available at https://github.com/Mamba413/CausalGraphCut.} }
Endnote
%0 Conference Paper %T Balancing Interference and Correlation in Spatial Experimental Designs: A Causal Graph Cut Approach %A Jin Zhu %A Jingyi Li %A Hongyi Zhou %A Yinan Lin %A Zhenhua Lin %A Chengchun Shi %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-zhu25l %I PMLR %P 79918--79937 %U https://proceedings.mlr.press/v267/zhu25l.html %V 267 %X This paper focuses on the design of spatial experiments to optimize the amount of information derived from the experimental data and enhance the accuracy of the resulting causal effect estimator. We propose a surrogate function for the mean squared error (MSE) of the estimator, which facilitates the use of classical graph cut algorithms to learn the optimal design. Our proposal offers three key advances: (1) it accommodates moderate to large spatial interference effects; (2) it adapts to different spatial covariance functions; (3) it is computationally efficient. Theoretical results and numerical experiments based on synthetic environments and a dispatch simulator that models a city-scale ridesharing market, further validate the effectiveness of our design. A python implementation of our method is available at https://github.com/Mamba413/CausalGraphCut.
APA
Zhu, J., Li, J., Zhou, H., Lin, Y., Lin, Z. & Shi, C.. (2025). Balancing Interference and Correlation in Spatial Experimental Designs: A Causal Graph Cut Approach. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:79918-79937 Available from https://proceedings.mlr.press/v267/zhu25l.html.

Related Material