David Blei

Causal inference from observational data is a vital problem, but it comes with strong assumptions. Most methods require that we observe all confounders, variables that correlate to both the causal variables (the treatment) and the effect of those...