Uncertainty Quantification in Probabilistic Machine Learning Models: Theory, Methods, and Insights

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Ajirak, Marzieh, Ravishankar, Anand, Djuric, Petar M.
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866915488289783808
author Ajirak, Marzieh
Ravishankar, Anand
Djuric, Petar M.
author_facet Ajirak, Marzieh
Ravishankar, Anand
Djuric, Petar M.
contents Uncertainty Quantification (UQ) is essential in probabilistic machine learning models, particularly for assessing the reliability of predictions. In this paper, we present a systematic framework for estimating both epistemic and aleatoric uncertainty in probabilistic models. We focus on Gaussian Process Latent Variable Models and employ scalable Random Fourier Features-based Gaussian Processes to approximate predictive distributions efficiently. We derive a theoretical formulation for UQ, propose a Monte Carlo sampling-based estimation method, and conduct experiments to evaluate the impact of uncertainty estimation. Our results provide insights into the sources of predictive uncertainty and illustrate the effectiveness of our approach in quantifying the confidence in the predictions.
format Preprint
id arxiv_https___arxiv_org_abs_2509_05877
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Uncertainty Quantification in Probabilistic Machine Learning Models: Theory, Methods, and Insights
Ajirak, Marzieh
Ravishankar, Anand
Djuric, Petar M.
Machine Learning
Artificial Intelligence
Uncertainty Quantification (UQ) is essential in probabilistic machine learning models, particularly for assessing the reliability of predictions. In this paper, we present a systematic framework for estimating both epistemic and aleatoric uncertainty in probabilistic models. We focus on Gaussian Process Latent Variable Models and employ scalable Random Fourier Features-based Gaussian Processes to approximate predictive distributions efficiently. We derive a theoretical formulation for UQ, propose a Monte Carlo sampling-based estimation method, and conduct experiments to evaluate the impact of uncertainty estimation. Our results provide insights into the sources of predictive uncertainty and illustrate the effectiveness of our approach in quantifying the confidence in the predictions.
title Uncertainty Quantification in Probabilistic Machine Learning Models: Theory, Methods, and Insights
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2509.05877