Rank-Learner : Orthogonal Ranking of Treatment Effects
Henri Arno, Dennis Frauen, Emil Javurek, Thomas Demeester, Stefan Feuerriegel · 2026
Read Rank-Learner : Orthogonal Ranking of Treatment Effects by Henri Arno, Dennis Frauen, Emil Javurek, Thomas Demeester, Stefan Feuerriegel (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-03517.txt.
Description
Many decisions need a ranking, not a number: which patients get preventive care first, or which customers an ad would sway most. Henri Arno and 4 co-authors introduce Rank-Learner in a 2026 paper. The two-stage method learns the ordering of treatment effects directly from observational data. Estimating each effect precisely first, they show, solves a harder problem than ranking requires.
Rank-Learner optimizes a pairwise objective that recovers the true ordering without ever estimating individual effects. The design is Neyman-orthogonal, giving theoretical protection against errors in the background nuisance estimates. It also works with any machine learning model under the hood. In extensive experiments it consistently beats standard effect estimators and non-orthogonal ranking methods.
Catalog record
| CREATOR | Henri Arno, Dennis Frauen, Emil Javurek, Thomas Demeester, Stefan Feuerriegel |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 105 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-03517