License: 
Creative Commons Attribution 4.0 International license (CC BY 4.0)
When quoting this document, please refer to the following
DOI: 10.4230/LIPIcs.ITCS.2023.92
URN: urn:nbn:de:0030-drops-175951
URL: http://dagstuhl.sunsite.rwth-aachen.de/volltexte/2023/17595/
Rothblum, Guy N. ;
Yona, Gal
Decision-Making Under Miscalibration
Abstract
How should we use ML-based predictions (e.g., risk of heart attack) to inform downstream binary classification decisions (e.g., undergoing a medical procedure)? When the risk estimates are perfectly calibrated, the answer is well understood: a classification problem’s cost structure induces an optimal treatment threshold j^⋆. In practice, however, predictors are often miscalibrated, and this can lead to harmful decisions. This raises a fundamental question: how should one use potentially miscalibrated predictions to inform binary decisions?
In this work, we study this question from the perspective of algorithmic fairness. Specifically, we focus on the impact of decisions on protected demographic subgroups, when we are only given a bound on the predictor’s anticipated degree of subgroup-miscalibration. We formalize a natural (distribution-free) solution concept for translating predictions into decisions: given anticipated miscalibration of α, we propose using the threshold j that minimizes the worst-case regret over all α-miscalibrated predictors, where the regret is the difference in clinical utility between using the threshold in question and using the optimal threshold in hindsight. We provide closed form expressions for j when miscalibration is measured using both expected and maximum calibration error which reveal that it indeed differs from j^⋆ (the optimal threshold under perfect calibration).
BibTeX - Entry
@InProceedings{rothblum_et_al:LIPIcs.ITCS.2023.92,
author = {Rothblum, Guy N. and Yona, Gal},
title = {{Decision-Making Under Miscalibration}},
booktitle = {14th Innovations in Theoretical Computer Science Conference (ITCS 2023)},
pages = {92:1--92:20},
series = {Leibniz International Proceedings in Informatics (LIPIcs)},
ISBN = {978-3-95977-263-1},
ISSN = {1868-8969},
year = {2023},
volume = {251},
editor = {Tauman Kalai, Yael},
publisher = {Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
address = {Dagstuhl, Germany},
URL = {https://drops.dagstuhl.de/opus/volltexte/2023/17595},
URN = {urn:nbn:de:0030-drops-175951},
doi = {10.4230/LIPIcs.ITCS.2023.92},
annote = {Keywords: risk prediction, calibration, algorithmic fairness, multi-group fairness}
}
|
Keywords: |
|
risk prediction, calibration, algorithmic fairness, multi-group fairness |
|
Collection: |
|
14th Innovations in Theoretical Computer Science Conference (ITCS 2023) |
|
Issue Date: |
|
2023 |
|
Date of publication: |
|
01.02.2023 |
