Conformal Multi-Target Hyperrectangles

Max Sampson; Kung-Sik Chan

doi:10.1002/sam.11710

Back

Journal article

Open access

Peer reviewed

Conformal Multi-Target Hyperrectangles

Max Sampson and Kung-Sik Chan

Statistical analysis and data mining, Vol.17(5), e11710

10/01/2024

DOI: 10.1002/sam.11710

Appears in UI Libraries Support Open Access

Files and links (1)

url

https://doi.org/10.1002/sam.11710View

Published (Version of record) Open Access

Abstract

We propose conformal hyperrectangular prediction regions for multi-target regression. We propose split conformal prediction algorithms for both point and quantile regression to form hyperrectangular prediction regions, which allow for easy marginal interpretation and do not require covariance estimation. In practice, it is preferable that a prediction region is balanced, that is, having identical marginal prediction coverage, since prediction accuracy is generally equally important across components of the response vector. The proposed algorithms possess two desirable properties, namely, tight asymptotic overall nominal coverage as well as asymptotic balance, that is, identical asymptotic marginal coverage, under mild conditions. We then compare our methods to some existing methods on both simulated and real data sets. Our simulation results and real data analysis show that our methods outperform existing methods while achieving the desired nominal coverage and good balance between dimensions.

Computer Science

Mathematics

Physical Sciences

Technology

balance

conformal prediction

multi-output prediction

multi-target regression

uncertainty quantification

UIOWA OA Agreement

Details

Title: Subtitle: Conformal Multi-Target Hyperrectangles
Creators: Max Sampson - University of Iowa
Kung-Sik Chan - University of Iowa
Resource Type: Journal article
Publication Details: Statistical analysis and data mining, Vol.17(5), e11710
Publisher: Wiley
DOI: 10.1002/sam.11710
ISSN: 1932-1864
eISSN: 1932-1872
Number of pages: 16
Grant note: National Institutes of Health Predoctoral Training Grant; United States Department of Health & Human Services; National Institutes of Health (NIH) - USA T32 HL 144461 / National Institutes of Health Predoctoral Training; United States Department of Health & Human Services; National Institutes of Health (NIH) - USA
Language: English
Date published: 10/01/2024
Academic Unit: Statistics and Actuarial Science; Radiology
Record Identifier: 9984719265602771

Metrics

1 Record Views