Exact Minimum-Volume Confidence Set Intersection for Multinomial Outcomes 1 The Scripps Research Institute, La Jolla, CA, USA. 2 Brown University, Providence, RI, USA. 3 University of Wisconsin–Madison, Madison, WI, USA.

Heguang Lin 1, Binhao Chen 2, Mengze Li 3, Daniel Pimentel-Alarcón 3, Matthew L. Malloy 3 · 2026

Read Exact Minimum-Volume Confidence Set Intersection for Multinomial Outcomes 1 The Scripps Research Institute, La Jolla, CA, USA. 2 Brown University, Providence, RI, USA. 3 University of Wisconsin–Madison, Madison, WI, USA. by Heguang Lin 1, Binhao Chen 2, Mengze Li 3, Daniel Pimentel-Alarcón 3, Matthew L. Malloy 3 (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-2601-18145.txt.

Description

Heguang Lin and co-authors compute exact confidence sets for multinomial outcomes. Their 2026 preprint deals with a basic tool of data science. Confidence sets underpin A/B testing and the analysis of learning algorithms. For outcomes with several categories, a good set should be as small as truth allows. The team studies the intersection of confidence sets with minimum volume, and they make the computation exact.

Exactness matters when sets guide product or policy choices. A set that is too wide wastes data; one that cheats breaks trust. Minimum volume squeezes the most from each sample. The work serves testers who must state what the data really support.

Catalog record

CREATORHeguang Lin 1, Binhao Chen 2, Mengze Li 3, Daniel Pimentel-Alarcón 3, Matthew L. Malloy 3
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE55 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-18145

Browse the full library · About the Tower of Babel