Inference for Clustering: Conformal Sets for Cluster Labels

YoonHaeng HurDepartment of Statistics, Columbia Universityand, Anirban NathDepartment of Statistics, Columbia UniversityandGenevera AllenDepartment of Statistics, Columbia University · 2026

Read Inference for Clustering: Conformal Sets for Cluster Labels by YoonHaeng HurDepartment of Statistics, Columbia Universityand, Anirban NathDepartment of Statistics, Columbia UniversityandGenevera AllenDepartment of Statistics, Columbia University (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-2604-03488.txt.

Description

Hur, Nath, and Allen's 2026 method returns trust sets for cluster labels, a quantity rarely given strict guarantees. Their split conformal clustering with random labels draws labels from soft cluster labels and fits a soft sorter to them, and tunes scores to build trust sets at any query point. Tests show target coverage with useful set sizes, and a use on single-cell RNA-seq cell types shows practical value.

Catalog record

CREATORYoonHaeng HurDepartment of Statistics, Columbia Universityand, Anirban NathDepartment of Statistics, Columbia UniversityandGenevera AllenDepartment of Statistics, Columbia University
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE87899
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.03488

Browse the full library · About the Tower of Babel