A production-grade pipeline for causal effect estimation on large-scale user interaction data. Implements difference-in-differences, propensity score matching, and automated power analysis on 500K+ ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results