Estimating Direct and Indirect Causal Effects of Spatiotemporal Interventions in Presence of Spatial Interference
This addresses a key challenge in causal inference for spatially correlated data, such as in climate science, by providing a novel method to quantify direct and indirect effects, though it is incremental in applying deep learning to this specific bottleneck.
The paper tackles the problem of estimating causal effects in spatiotemporal settings with spatial interference, where treatments at one location affect outcomes elsewhere, by proposing a deep learning-based potential outcome model. It shows advantages over baseline methods on synthetic datasets and aligns with domain knowledge on a real-world climate dataset.
Spatial interference (SI) occurs when the treatment at one location affects the outcomes at other locations. Accounting for spatial interference in spatiotemporal settings poses further challenges as interference violates the stable unit treatment value assumption, making it infeasible for standard causal inference methods to quantify the effects of time-varying treatment at spatially varying outcomes. In this paper, we first formalize the concept of spatial interference in case of time-varying treatment assignments by extending the potential outcome framework under the assumption of no unmeasured confounding. We then propose our deep learning based potential outcome model for spatiotemporal causal inference. We utilize latent factor modeling to reduce the bias due to time-varying confounding while leveraging the power of U-Net architecture to capture global and local spatial interference in data over time. Our causal estimators are an extension of average treatment effect (ATE) for estimating direct (DATE) and indirect effects (IATE) of spatial interference on treated and untreated data. Being the first of its kind deep learning based spatiotemporal causal inference technique, our approach shows advantages over several baseline methods based on the experiment results on two synthetic datasets, with and without spatial interference. Our results on real-world climate dataset also align with domain knowledge, further demonstrating the effectiveness of our proposed method.