Principled Approaches for Learning to Defer with Multiple Experts

Anqi Mao, Mehryar Mohri, Yutao Zhong

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We present a study of surrogate losses and algorithms for the general problem of learning to defer with multiple experts. We first introduce a new family of surrogate losses specifically tailored for the multiple-expert setting, where the prediction and deferral functions are learned simultaneously. We then prove that these surrogate losses benefit from strong H-consistency bounds. We illustrate the application of our analysis through several examples of practical surrogate losses, for which we give explicit guarantees. These loss functions readily lead to the design of new learning to defer algorithms based on their minimization. While the main focus of this work is a theoretical analysis, we also report the results of several experiments on SVHN and CIFAR-10 datasets.

Original languageEnglish (US)
Title of host publicationArtificial Intelligence and Image Analysis - 18th International Symposium on Artificial Intelligence and Mathematics, ISAIM 2024, and 22nd International Workshop on Combinatorial Image Analysis, IWCIA 2024, Revised Selected Papers
EditorsReneta P. Barneva, Valentin E. Brimkov, Valentin E. Brimkov, Claudio Gentile, Aldo Pacchiano
PublisherSpringer Science and Business Media Deutschland GmbH
Pages107-135
Number of pages29
ISBN (Print)9783031637346
DOIs
StatePublished - 2024
Event18th International Symposium on Artificial Intelligence and Mathematics, ISAIM 2024, and 22nd International Workshop on Combinatorial Image Analysis, IWCIA 2024 - Fort Lauderdale, United States
Duration: Jan 8 2024Jan 10 2024

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume14494 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference18th International Symposium on Artificial Intelligence and Mathematics, ISAIM 2024, and 22nd International Workshop on Combinatorial Image Analysis, IWCIA 2024
Country/TerritoryUnited States
CityFort Lauderdale
Period1/8/241/10/24

Keywords

  • Consistency
  • Learning theory
  • Learning to defer

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science

Fingerprint

Dive into the research topics of 'Principled Approaches for Learning to Defer with Multiple Experts'. Together they form a unique fingerprint.

Cite this