Saved in:
Bibliographic Details
Main Authors: Kuang, Qi, Gang, Bowen, Xia, Yin
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2512.01423
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910111322079232
author Kuang, Qi
Gang, Bowen
Xia, Yin
author_facet Kuang, Qi
Gang, Bowen
Xia, Yin
contents In large-scale hypothesis testing, computing exact $p$-values or $e$-values is often resource-intensive, creating a need for budget-aware inferential methods. We propose a general framework for active hypothesis testing that leverages inexpensive auxiliary statistics to allocate a global computational budget. For each hypothesis, our data-adaptive procedure probabilistically decides whether to compute the exact test statistic or a transformed proxy, guaranteeing a valid $p$-value or $e$-value while satisfying the exact budget constraint. Theoretical guarantees are established for our constructions, showing that the procedure achieves optimality for $e$-values and for $p$-values under independence, and admissibility for $p$-values under general dependence. Empirical results from simulations and two real-world applications, including a large-scale genome-wide association study (GWAS) and a clinical prediction task leveraging large language models (LLM), demonstrate that our framework improves statistical efficiency under fixed resource limits.
format Preprint
id arxiv_https___arxiv_org_abs_2512_01423
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Active Hypothesis Testing under Computational Budgets with Applications to GWAS and LLM
Kuang, Qi
Gang, Bowen
Xia, Yin
Methodology
In large-scale hypothesis testing, computing exact $p$-values or $e$-values is often resource-intensive, creating a need for budget-aware inferential methods. We propose a general framework for active hypothesis testing that leverages inexpensive auxiliary statistics to allocate a global computational budget. For each hypothesis, our data-adaptive procedure probabilistically decides whether to compute the exact test statistic or a transformed proxy, guaranteeing a valid $p$-value or $e$-value while satisfying the exact budget constraint. Theoretical guarantees are established for our constructions, showing that the procedure achieves optimality for $e$-values and for $p$-values under independence, and admissibility for $p$-values under general dependence. Empirical results from simulations and two real-world applications, including a large-scale genome-wide association study (GWAS) and a clinical prediction task leveraging large language models (LLM), demonstrate that our framework improves statistical efficiency under fixed resource limits.
title Active Hypothesis Testing under Computational Budgets with Applications to GWAS and LLM
topic Methodology
url https://arxiv.org/abs/2512.01423