Abstract: There has been a recent surge in attention towards trustworthy AI, especially as it starts playing a pivotal role in high-stakes domains such as healthcare, the justice system, and finance. Causal inference emerges as a promising path toward building AI systems that are stable, fair, and explainable. However, it often hinges on precise and strong assumptions. In this talk, I introduce sparse causal learning as a common ground between trustworthy AI and robust causal inference. Here, sparsity plays a dual role in enhancing explainability and ensuring the robust identification of causal effects. Specifically, I reconsider the supervised learning problem of predicting an outcome using multiple predictors through the lens of causality. I demonstrate that it is possible to remove spurious correlations caused by unmeasured confounding by leveraging low-dimensional structures in the predictors. I study its identifiability using an expert voting approach and show that sparsity provides a promising path to transforming exact causal inference methods into multiply robust identification frameworks. Furthermore, I introduce the synthetic instrument, a novel tool for constructing instrumental variables and estimating causal effects. This new approach leads to algorithms that are theoretically justifiable, computationally feasible, and statistically sound.

https://sites.google.com/site/linbowangpku/