EUVIMEDEUVIMED entdecken
Open Access

Prediction sets adaptive to unknown covariate shift

Hongxiang Qiu, Edgar Dobriban, Eric Tchetgen Tchetgen

Journal of the Royal Statistical Society Series B: Statistical Methodology · 02.02.2024 · 85 (5) · 1680-1705

DOI: 10.1093/jrsssb/qkad069 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of the Royal Statistical Society Series B: Statistical Methodology, online erschienen am 17.07.2023. Verfasst von Hongxiang Qiu und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Statistics, The Wharton School, University of Pennsylvania, Philadelphia, PA 19104, USA. Die Arbeit wurde bisher 13-mal zitiert (Crossref) und nennt 84 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: NIA NIH HHS, NCI NIH HHS, NIGMS NIH HHS.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Predicting sets of outcomes—instead of unique outcomes—is a promising solution to uncertainty quantification in statistical learning. Despite a rich literature on constructing prediction sets with statistical guarantees, adapting to unknown covariate shift—a prevalent issue in practice—poses a serious unsolved challenge. In this article, we show that prediction sets with finite-sample coverage guarantee are uninformative and propose a novel flexible distribution-free method, PredSet-1Step, to efficiently construct prediction sets with an asymptotic coverage guarantee under unknown covariate shift. We formally show that our method is asymptotically probably approximately correct, having well-calibrated coverage error with high confidence for large samples. We illustrate that it achieves nominal coverage in a number of experiments and a data set concerning HIV risk prediction in a South African cohort study. Our theory hinges on a new bound for the convergence rate of the coverage of Wald confidence intervals based on general asymptotically linear estimators.

Schlagwörter

Stichwörter der Autor:innen

PAC guaranteecovariate shiftmachine learningnonparametric inferencenonparametric modelprediction set

Autor:innen und Institutionen

  • Department of Statistics, The Wharton School, University of Pennsylvania, Philadelphia, PA 19104, USA.

Förderung

  • NIA NIH HHS
  • NCI NIH HHS
  • NIGMS NIH HHS
  • DMS
  • NIH

Zitieren

APA

Qiu, H., Dobriban, E., & Tchetgen Tchetgen, E. (2024). Prediction sets adaptive to unknown covariate shift. Journal of the Royal Statistical Society Series B: Statistical Methodology, 85(5), 1680-1705. https://doi.org/10.1093/jrsssb/qkad069

Vancouver

Qiu H, Dobriban E, Tchetgen Tchetgen E. Prediction sets adaptive to unknown covariate shift. Journal of the Royal Statistical Society Series B: Statistical Methodology. 2024;85(5):1680-1705. doi:10.1093/jrsssb/qkad069

BibTeX
@article{Qiu2024,
  author = {Qiu, Hongxiang and Dobriban, Edgar and Tchetgen Tchetgen, Eric},
  title = {Prediction sets adaptive to unknown covariate shift},
  journal = {Journal of the Royal Statistical Society Series B: Statistical Methodology},
  year = {2024},
  volume = {85},
  number = {5},
  pages = {1680--1705},
  doi = {10.1093/jrsssb/qkad069}
}
Kostenlos starten →