cs.LG · 2608.07630 Copy arXiv ID · Aug 7, 2026 Save Tracing sources of epistemic uncertainty in deep learning predictions: homo- and hetero-scedastic linearized estimators Authors: Pierre Nodet , Thomas George
Organizations: Orange Research, Châtillon, France
Abstract We adapt two classical statistical estimators for quantifying uncertainty to modern deep learning, in order to provide clearer insights into uncertainty attributable to two sources : aleatoric uncertainty, or locally scarce data. Our approach leverages recent advances in approximate Fisher Information Matrices, to enable scaling to actual architectures. Experimental results demonstrate how each test points is differentially impacted by both sources, highlighting the practical utility of our estimators in improving the robustness of real-world applications.
Explore similar work May 1, 2026 · Yao Ni, Jeremie Houssineau, Yew-Soon Ong +1 Predictive Uncertainty Posterior
May 29, 2026 · Joseph Wilson, Chris van der Heide, Liam Hodgkinson +1 Reliable Uncertainty Quantification Logistic Regression
May 1, 2026 · cs.LG J/K move · Enter open · S save
Yao Ni, Jeremie Houssineau, Yew-Soon Ong, Piotr Koniusz
Nanyang Technological University · A*STAR · The University of New South Wales · Data61 CSIRO
Deep neural networks achieve impressive results across diverse applications, yet their overconfidence on unseen inputs necessitates reliable epistemic uncertainty modeling. Existing methods for uncertainty modeling face a fundamental dilemma: Bayesian approaches provide principled estimates but remain computationally prohibitive, while efficient second-order predictors lack rigorous connections between their specific objectives and epistemic uncertainty quantification. To resolve this dilemma, we introduce Dirichlet-approximated possibilistic posterior predictions (DAPPr), a principled framework grounded in possibility theory. We define a possibilistic posterior over parameters, project it to the prediction space via supremum operators, and approximate the projected posterior using learnable Dirichlet possibility functions. This projection-and-approximation strategy yields a simple training objective with closed-form solutions. Despite its simplicity, extensive experiments across diverse benchmarks show that DAPPr achieves competitive or superior uncertainty quantification performance over state-of-the-art second-order predictors while maintaining both principled derivation and computational efficiency. Code is available at https://github.com/MaxwellYaoNi/DAPPr.