Beyond generalization: a theory of robustness in machine learning.
The term robustness is ubiquitous in modern Machine Learning (ML). However, its meaning varies depending on context and community. Researchers either focus on narrow technical definitions, such as adversarial robustness, natural distribution shifts, and performativity, or they simply leave open what...
| Publicado en: | Synthese Vol. 202; no. 4; pp. 1 - 29 |
|---|---|
| Autores principales: | , |
| Formato: | Artículo |
| Publicado: |
Springer Nature
Oct2023
|
| Acceso en línea: | Ver este registro en EBSCOhost |
| fields | @attributes: recordID: 1 pdfLink: plink: https://search.ebscohost.com/login.aspx?direct=true&db=hlh&AN=172379025&site=ehost-live header: @attributes: shortDbName: hlh uiTerm: 172379025 longDbName: Humanities International Complete uiTag: AN controlInfo: bkinfo: jinfo: jid: 00397857 4LI jtl: Synthese issn: 00397857 maglogo: N pubinfo: dt: Oct2023 vid: 202 iid: 4 pid: 237 pub: Springer Nature artinfo: ui: 172379025 10.1007/s11229-023-04334-9 ppf: 1 ppct: 28 formats: fmt: – @attributes: type: T – @attributes: type: P size: 709KB tig: atl: Beyond generalization: a theory of robustness in machine learning. aug: au: Freiesleben, Timo Grote, Thomas affil: https://ror.org/03a1kwz48 Cluster of Excellence: "Machine Learning: New Perspectives for Science", University of Tübingen, Maria-von-Linden-Straße 6, 72076, Tübingen, Germany sug: keyword: Extrapolation Generalization Machine Learning Models in Science Robustness Uncertainty ab: The term robustness is ubiquitous in modern Machine Learning (ML). However, its meaning varies depending on context and community. Researchers either focus on narrow technical definitions, such as adversarial robustness, natural distribution shifts, and performativity, or they simply leave open what exactly they mean by robustness. In this paper, we provide a conceptual analysis of the term robustness, with the aim to develop a common language, that allows us to weave together different strands of robustness research. We define robustness as the relative stability of a robustness target with respect to specific interventions on a modifier. Our account captures the various sub-types of robustness that are discussed in the research literature, including robustness to distribution shifts, prediction robustness, or the robustness of algorithmic explanations. Finally, we delineate robustness from adjacent key concepts in ML, such as extrapolation, generalization, and uncertainty, and establish it as an independent epistemic concept. pubtype: Academic Journal doctype: Article src: R language: English refInfo: copyright: @attributes: flag: Y custom: Synthese is a copyright of Springer, 2023. All Rights Reserved. item: Synthese holder: Springer Nature dt: @attributes: year: 2023 holdings: @attributes: islocal: N |
|---|