Reproduce “Statistical Learning from Attribution Sets” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Lorne Applebaum, Robert Busa-Fekete, August Y. Chen, Claudio Gentile, Tomer Koren, Aryan Mokhtari
The authors develop a method for learning click-level conversion predictions from privacy-preserving attribution sets, which identify candidate clicks but not the converting click. Using known probabilities for the true click’s position, the method estimates prediction loss without assigning a guessed label. A formal guarantee applies under an oblivious, feature-independent model; preliminary modified-dataset experiments outperform random- and maximum-prior-label heuristics, not real-traffic deployment.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.