FROM RISK TO UNCERTAINTY: GENERATING PREDICTIVE UNCERTAINTY MEASURES VIA BAYESIAN ESTIMATION

Nikita Kotelevskii; Vladimir Kondratyev; Martin Takáč; Éric Moulines; Maxim Panov

FROM RISK TO UNCERTAINTY: GENERATING PREDICTIVE UNCERTAINTY MEASURES VIA BAYESIAN ESTIMATION

Nikita Kotelevskii
, Vladimir Kondratyev
, Martin Takáč
, Éric Moulines
, Maxim Panov

Mohamed bin Zayed University of Artificial Intelligence (MBZUAI)
CAIT
École Polytechnique

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

6 Citations (Scopus)

Abstract

There are various measures of predictive uncertainty in the literature, but their relationships to each other remain unclear. This paper uses a decomposition of statistical pointwise risk into components, associated with different sources of predictive uncertainty, namely aleatoric uncertainty (inherent data variability) and epistemic uncertainty (model-related uncertainty). Together with Bayesian methods, applied as an approximation, we build a framework that allows one to generate different predictive uncertainty measures. We validate our method on image datasets by evaluating its performance in detecting out-of-distribution and misclassified instances using the AUROC metric. The experimental results confirm that the measures derived from our framework are useful for the considered downstream tasks.

Original language	English
Title of host publication	13th International Conference on Learning Representations, ICLR 2025
Publisher	International Conference on Learning Representations, ICLR
Pages	27157-27193
Number of pages	37
ISBN (Electronic)	9798331320850
Publication status	Published - 1 Jan 2025
Externally published	Yes
Event	13th International Conference on Learning Representations, ICLR 2025 - Singapore, Singapore Duration: 24 Apr 2025 → 28 Apr 2025

Publication series

Name	13th International Conference on Learning Representations, ICLR 2025

Conference

Conference	13th International Conference on Learning Representations, ICLR 2025
Country/Territory	Singapore
City	Singapore
Period	24/04/25 → 28/04/25

Cite this

Kotelevskii, N., Kondratyev, V., Takáč, M., Moulines, É., & Panov, M. (2025). FROM RISK TO UNCERTAINTY: GENERATING PREDICTIVE UNCERTAINTY MEASURES VIA BAYESIAN ESTIMATION. In 13th International Conference on Learning Representations, ICLR 2025 (pp. 27157-27193). (13th International Conference on Learning Representations, ICLR 2025). International Conference on Learning Representations, ICLR.

Kotelevskii, Nikita ; Kondratyev, Vladimir ; Takáč, Martin et al. / FROM RISK TO UNCERTAINTY : GENERATING PREDICTIVE UNCERTAINTY MEASURES VIA BAYESIAN ESTIMATION. 13th International Conference on Learning Representations, ICLR 2025. International Conference on Learning Representations, ICLR, 2025. pp. 27157-27193 (13th International Conference on Learning Representations, ICLR 2025).

@inproceedings{4c759cc69b4847b88af2b0e20b6052a3,

title = "FROM RISK TO UNCERTAINTY: GENERATING PREDICTIVE UNCERTAINTY MEASURES VIA BAYESIAN ESTIMATION",

abstract = "There are various measures of predictive uncertainty in the literature, but their relationships to each other remain unclear. This paper uses a decomposition of statistical pointwise risk into components, associated with different sources of predictive uncertainty, namely aleatoric uncertainty (inherent data variability) and epistemic uncertainty (model-related uncertainty). Together with Bayesian methods, applied as an approximation, we build a framework that allows one to generate different predictive uncertainty measures. We validate our method on image datasets by evaluating its performance in detecting out-of-distribution and misclassified instances using the AUROC metric. The experimental results confirm that the measures derived from our framework are useful for the considered downstream tasks.",

author = "Nikita Kotelevskii and Vladimir Kondratyev and Martin Tak{\'a}{\v c} and {\'E}ric Moulines and Maxim Panov",

note = "Publisher Copyright: {\textcopyright} 2025 13th International Conference on Learning Representations, ICLR 2025. All rights reserved.; 13th International Conference on Learning Representations, ICLR 2025 ; Conference date: 24-04-2025 Through 28-04-2025",

year = "2025",

month = jan,

day = "1",

language = "English",

series = "13th International Conference on Learning Representations, ICLR 2025",

publisher = "International Conference on Learning Representations, ICLR",

pages = "27157--27193",

booktitle = "13th International Conference on Learning Representations, ICLR 2025",

}

Kotelevskii, N, Kondratyev, V, Takáč, M, Moulines, É & Panov, M 2025, FROM RISK TO UNCERTAINTY: GENERATING PREDICTIVE UNCERTAINTY MEASURES VIA BAYESIAN ESTIMATION. in 13th International Conference on Learning Representations, ICLR 2025. 13th International Conference on Learning Representations, ICLR 2025, International Conference on Learning Representations, ICLR, pp. 27157-27193, 13th International Conference on Learning Representations, ICLR 2025, Singapore, Singapore, 24/04/25.

FROM RISK TO UNCERTAINTY: GENERATING PREDICTIVE UNCERTAINTY MEASURES VIA BAYESIAN ESTIMATION. / Kotelevskii, Nikita; Kondratyev, Vladimir; Takáč, Martin et al.
13th International Conference on Learning Representations, ICLR 2025. International Conference on Learning Representations, ICLR, 2025. p. 27157-27193 (13th International Conference on Learning Representations, ICLR 2025).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - FROM RISK TO UNCERTAINTY

T2 - 13th International Conference on Learning Representations, ICLR 2025

AU - Kotelevskii, Nikita

AU - Kondratyev, Vladimir

AU - Takáč, Martin

AU - Moulines, Éric

AU - Panov, Maxim

PY - 2025/1/1

Y1 - 2025/1/1

N2 - There are various measures of predictive uncertainty in the literature, but their relationships to each other remain unclear. This paper uses a decomposition of statistical pointwise risk into components, associated with different sources of predictive uncertainty, namely aleatoric uncertainty (inherent data variability) and epistemic uncertainty (model-related uncertainty). Together with Bayesian methods, applied as an approximation, we build a framework that allows one to generate different predictive uncertainty measures. We validate our method on image datasets by evaluating its performance in detecting out-of-distribution and misclassified instances using the AUROC metric. The experimental results confirm that the measures derived from our framework are useful for the considered downstream tasks.

AB - There are various measures of predictive uncertainty in the literature, but their relationships to each other remain unclear. This paper uses a decomposition of statistical pointwise risk into components, associated with different sources of predictive uncertainty, namely aleatoric uncertainty (inherent data variability) and epistemic uncertainty (model-related uncertainty). Together with Bayesian methods, applied as an approximation, we build a framework that allows one to generate different predictive uncertainty measures. We validate our method on image datasets by evaluating its performance in detecting out-of-distribution and misclassified instances using the AUROC metric. The experimental results confirm that the measures derived from our framework are useful for the considered downstream tasks.

UR - https://www.scopus.com/pages/publications/105010224483

M3 - Conference contribution

AN - SCOPUS:105010224483

T3 - 13th International Conference on Learning Representations, ICLR 2025

SP - 27157

EP - 27193

BT - 13th International Conference on Learning Representations, ICLR 2025

PB - International Conference on Learning Representations, ICLR

Y2 - 24 April 2025 through 28 April 2025

ER -

Kotelevskii N, Kondratyev V, Takáč M, Moulines É, Panov M. FROM RISK TO UNCERTAINTY: GENERATING PREDICTIVE UNCERTAINTY MEASURES VIA BAYESIAN ESTIMATION. In 13th International Conference on Learning Representations, ICLR 2025. International Conference on Learning Representations, ICLR. 2025. p. 27157-27193. (13th International Conference on Learning Representations, ICLR 2025).

FROM RISK TO UNCERTAINTY: GENERATING PREDICTIVE UNCERTAINTY MEASURES VIA BAYESIAN ESTIMATION

Abstract

Publication series

Conference

Other files and links

Fingerprint

Cite this