From Risk to Uncertainty: Generating Predictive Uncertainty Measures via Bayesian Estimation

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Kotelevskii, Nikita, Kondratyev, Vladimir, Takáč, Martin, Moulines, Éric, Panov, Maxim
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866929716313718784
author Kotelevskii, Nikita
Kondratyev, Vladimir
Takáč, Martin
Moulines, Éric
Panov, Maxim
author_facet Kotelevskii, Nikita
Kondratyev, Vladimir
Takáč, Martin
Moulines, Éric
Panov, Maxim
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2402_10727
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle From Risk to Uncertainty: Generating Predictive Uncertainty Measures via Bayesian Estimation
Kotelevskii, Nikita
Kondratyev, Vladimir
Takáč, Martin
Moulines, Éric
Panov, Maxim
Machine Learning
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.
title From Risk to Uncertainty: Generating Predictive Uncertainty Measures via Bayesian Estimation
topic Machine Learning
url https://arxiv.org/abs/2402.10727