mlsanityData to 13 Oct 2026

Epistemic Deep Learning

arXiv 2206.07609 ↗First posted 15 Jun 2022cs.LG cs.NE stat.MLcited by 15

Not on the map. Why?

Abstract (arXiv)

The belief function approach to uncertainty quantification as proposed in the Demspter-Shafer theory of evidence is established upon the general mathematical models for set-valued observations, called random sets. Set-valued predictions are the most natural representations of uncertainty in machine learning. In this paper, we introduce a concept called epistemic deep learning based on the random-set interpretation of belief functions to model epistemic learning in deep neural networks. We propose a novel random-set convolutional neural network for classification that produces scores for sets of classes by learning set-valued ground truth representations. We evaluate different formulations of entropy and distance measures for belief functions as viable loss functions for these random-set networks. We also discuss methods for evaluating the quality of epistemic predictions and the performance of epistemic random-set neural networks. We demonstrate through experiments that the epistemic approach produces better performance results when compared to traditional approaches of estimating uncertainty.

Key facts. Cited by 15 papers in the map (data through 13 Oct 2026). Builds on 11 of its 100 references that are in the map. First posted on arXiv 15 Jun 2022.

Builds on

11 of 100 references.

References · 11

  1. 2017What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?cited by 2,340
  2. 2019Aleatoric and Epistemic Uncertainty in Machine Learning: An Introduction to Concepts and Methodscited by 910
  3. 2017Selective Classification for Deep Neural Networkscited by 825
  4. 2015The Limitations of Deep Learning in Adversarial Settingscited by 790
  5. 2018Evidential Deep Learning to Quantify Classification Uncertaintycited by 729
  6. 2020Hands-on Bayesian Neural Networks -- a Tutorial for Deep Learning Userscited by 337
  7. 2015Bayesian Convolutional Neural Networks with Bernoulli Approximate Variational Inferencecited by 159
  8. 2013Learning by Transductioncited by 120
  9. 2021Evaluating Adversarial Attacks on Driving Safety in Vision-Based Autonomous Vehiclescited by 24
  10. 2021An evidential classifier based on Dempster-Shafer theory and deep learningcited by 18
  11. 2013The Transferable Belief Model and Other Interpretations of Dempster-Shafer's Modelcited by 2

Built on by

Cited by year (2026 to 13 Oct)

Citing papers per year: 2023: 4; 2024: 2; 2025: 2; 2026: 7.

15 papers.

Sort:

By area:

  1. 2024CreINNs: Credal-Set Interval Neural Networks for Uncertainty Estimation in Classification Taskscited by 26
  2. 2023Random-Set Neural Networks (RS-NN)cited by 19
  3. 2025A Unified Evaluation Framework for Epistemic Predictionscited by 10
  4. 2023Reasoning with random sets: An agenda for the futurecited by 6
  5. 2024Generalising realisability in statistical learning theory under epistemic uncertaintycited by 5
  6. 2025Epistemic Wrapping for Uncertainty Quantificationcited by 5
  7. 2026Epistemic Generative Adversarial Networkscited by 4
  8. 2026Random-Set Graph Neural Networkscited by 3
  9. 2023A Hybrid Graph Network for Complex Activity Detection in Videocited by 2
  10. 2026Credal Large Language Models for Semantic Commitment under Uncertaintycited by 2
  11. 2026A neurosymbolic Approach with Epistemic Deep Learning for Hierarchical Image Classificationcited by 1
  12. 2026A formal definition and meta-model for a machine theory of mindcited by 1
3 more
  1. 2026What Intermediate Layers Know: Detecting Jailbreaks from Entropy Dynamicscited by 1
  2. 2023Spatiotemporal Event Graphs for Dynamic Scene Understandingcited by 0
  3. 2026XU-RS: Explaining Credal Width in Random-Set Language Modelscited by 0