no code implementations • 8 Dec 2023 • Björn Bokelmann, Stefan Lessmann
In our research, we show that heteroskedastictity in the training data can cause a bias of the uplift model ranking: individuals with the highest treatment effects can get accumulated in large numbers at the bottom of the ranking.
1 code implementation • 5 Oct 2022 • Björn Bokelmann, Stefan Lessmann
We theoretically analyze the variance of uplift evaluation metrics and derive possible methods of variance reduction, which are based on statistical adjustment of the outcome.