This talk is about identifiability of causal effects under unobserved confounding. We consider causal effects of a treatment on an outcome and rather than observing the confounder directly, we observe a single, potentially high-dimensional proxy variable of the confounder. Further, we assume that the mechanism generating the proxy is known. Under assumptions on this mechanism, we prove that causal effects can be recovered. Hence, we establish Single Proxy Identifiability of Causal Effects, or simply SPICE. Our results extend current research to higher dimensions, more flexible functional relationships, and a broader class of distributions. In this setting, we develop a neural network-based method to estimate causal effects. It can be applied, for example, in oncology, where a patient’s overall fitness acts as an unobserved confounder, influencing both the choice of treatment aimed at curing cancer and patient survival.
This is joint work with Sebastian Weichwald and Niklas Pfister.
