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

2025-05-26

AI Summary

This paper addresses the challenge of substantial bias in causal effect estimation under spatial experimental design, where both interference and spatial correlation coexist. We propose a novel graph-cut optimization framework for treatment assignment. Our key innovation is the first formulation of a mean squared error (MSE) surrogate objective, which recasts experimental allocation as a minimum-cut problem—enabling modeling of moderate-to-strong spatial interference and accommodating diverse covariance structures (e.g., exponential, Gaussian, Matérn). The method provides theoretical guarantees on optimality while ensuring computational scalability. Evaluated on synthetic data and city-scale ride-hailing dispatch simulations, our approach reduces MSE by up to 37% on average compared to state-of-the-art baselines, significantly improving causal estimation accuracy and efficiency of information utilization.

Technology Category

• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Reasoning under Uncertainty: Causality

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

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.

Problem

• Optimizing spatial experiments for accurate causal estimation
• Designing experiments with spatial interference and correlation
• Developing efficient graph cut algorithms for spatial designs

Innovation

• Uses graph cut algorithms for optimal design
• Adapts to various spatial covariance functions
• Computationally efficient for large interference effects