Statistical Learning from Attribution Sets
Lorne Applebaum, Robert Busa-Fekete, August Y. Chen, Claudio Gentile, Tomer Koren, Aryan Mokhtari · 2026
Read Statistical Learning from Attribution Sets by Lorne Applebaum, Robert Busa-Fekete, August Y. Chen, Claudio Gentile, Tomer Koren, Aryan Mokhtari (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-2602-06276.txt.
Description
Online ads face a privacy problem: browsers no longer say exactly which click caused a purchase. Lorne Applebaum and 5 co-authors study learning from attribution sets, where each conversion is linked only to a set of candidate clicks. Privacy-preserving browser tools report data in exactly this coarse form.
Working against an adversary that builds the sets from a prior guess, the team constructs an unbiased estimate of the true loss from these blurred signals. Standard risk minimization on that estimate comes with guarantees that grow stronger as the prior gets more informative, and stay robust when the prior is off. Simple tests on standard data beat the heuristics industry commonly uses, especially when sets are large or overlapping.
Catalog record
| CREATOR | Lorne Applebaum, Robert Busa-Fekete, August Y. Chen, Claudio Gentile, Tomer Koren, Aryan Mokhtari |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 123 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-06276