mlsanityData to 14 Oct 2026

Posterior Uncertainty Quantification in Neural Networks using Data Augmentation

arXiv 2403.12729 ↗First posted 18 Mar 2024stat.ML cs.LGcited by 9

Not on the map. Why?

Abstract (arXiv)

In this paper, we approach the problem of uncertainty quantification in deep learning through a predictive framework, which captures uncertainty in model parameters by specifying our assumptions about the predictive distribution of unseen future data. Under this view, we show that deep ensembling (Lakshminarayanan et al., 2017) is a fundamentally mis-specified model class, since it assumes that future data are supported on existing observations only -- a situation rarely encountered in practice. To address this limitation, we propose MixupMP, a method that constructs a more realistic predictive distribution using popular data augmentation techniques. MixupMP operates as a drop-in replacement for deep ensembles, where each ensemble member is trained on a random simulation from this predictive distribution. Grounded in the recently-proposed framework of Martingale posteriors (Fong et al., 2023), MixupMP returns samples from an implicitly defined Bayesian posterior. Our empirical analysis showcases that MixupMP achieves superior predictive performance and uncertainty quantification on various image classification datasets, when compared with existing Bayesian and non-Bayesian approaches.

Key facts. Cited by 9 papers in the map (data through 14 Oct 2026). Builds on 41 of its 74 references that are in the map. First posted on arXiv 18 Mar 2024.

Builds on

41 of 74 references.

References · 41

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  3. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,183
  4. 2015Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learningcited by 4,448
  5. 2017Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithmscited by 4,349
  6. 2016Simple and Scalable Predictive Uncertainty Estimation using Deep Ensemblescited by 3,926
  7. 2016Wide Residual Networkscited by 3,070
  8. 2019Can You Trust Your Model's Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shiftcited by 1,287
  9. 2016Deep Exploration via Bootstrapped DQNcited by 527
  10. 2020Bayesian Deep Learning and a Probabilistic Perspective of Generalizationcited by 480
  11. 2021A Survey of Data Augmentation Approaches for NLPcited by 409
  12. 2017Training Confidence-calibrated Classifiers for Detecting Out-of-Distribution Samplescited by 389
29 more
  1. 2021Laplace Redux -- Effortless Bayesian Deep Learningcited by 365
  2. 2014Stochastic Gradient Hamiltonian Monte Carlocited by 343
  3. 2019Gradient Descent Maximizes the Margin of Homogeneous Neural Networkscited by 320
  4. 2018Neural Processescited by 311
  5. 2021What Are Bayesian Neural Network Posteriors Really Like?cited by 308
  6. 2019On Mixup Training: Improved Calibration and Predictive Uncertainty for Deep Neural Networkscited by 305
  7. 2020Implicit Bias of Gradient Descent for Wide Two-layer Neural Networks Trained with the Logistic Losscited by 268
  8. 2018What is the Effect of Importance Weighting in Deep Learning?cited by 251
  9. 2018Randomized Prior Functions for Deep Reinforcement Learningcited by 219
  10. 2019Cyclical Stochastic Gradient MCMC for Bayesian Deep Learningcited by 203
  11. 2020Directional convergence and alignment in deep learningcited by 178
  12. 2018Understanding Measures of Uncertainty for Adversarial Example Detectioncited by 177
  13. 2020Uncertainty Quantification and Deep Ensemblescited by 139
  14. 2018Regularization Matters: Generalization and Optimization of Neural Nets v.s. their Induced Kernelcited by 131
  15. 2018Convergence of Gradient Descent on Separable Datacited by 110
  16. 2018Benchmarking Neural Network Robustness to Common Corruptions and Surface Variationscited by 99
  17. 2018Towards Safe Autonomous Driving: Capture Uncertainty in the Deep Neural Network For Lidar 3D Vehicle Detectioncited by 82
  18. 2022Deep Ensembles Work, But Are They Necessary?cited by 63
  19. 2019A Systematic Comparison of Bayesian Deep Learning Robustness in Diabetic Retinopathy Taskscited by 62
  20. 2020Combining Ensembles and Data Augmentation can Harm your Calibrationcited by 44
  21. 2023Tractable Function-Space Variational Inference in Bayesian Neural Networkscited by 41
  22. 2021MixSpeech: Data Augmentation for Low-resource Automatic Speech Recognitioncited by 39
  23. 2021Understanding the role of importance weighting for deep learningcited by 39
  24. 2021Is Importance Weighting Incompatible with Interpolating Classifiers?cited by 19
  25. 2018Nonparametric learning from Bayesian models with randomized objective functionscited by 17
  26. 2019Scalable Nonparametric Sampling from Multimodal Posteriors with the Posterior Bootstrapcited by 16
  27. 2023Martingale Posterior Neural Processescited by 16
  28. 2017Analysis of dropout learning regarded as ensemble learningcited by 15
  29. 2020Neural Bootstrappercited by 5

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 2; 2025: 5; 2026: 2.

9 papers.

  1. 2025How Confident are Video Models? Empowering Video Models to Express their Uncertaintycited by 6
  2. 2024Quasi-Bayes meets Vinescited by 5
  3. 2025Contextual Similarity Distillation: Ensemble Uncertainties with a Single Modelcited by 3
  4. 2025Efficient Post-Hoc Uncertainty Calibration via Variance-Based Smoothingcited by 3
  5. 2024Robust Classification by Coupling Data Mollification with Label Smoothingcited by 2
  6. 2025Understanding the Trade-offs in Accuracy and Uncertainty Quantification: Architecture and Inference Choices in Bayesian Neural Networkscited by 1
  7. 2025Optimizing Data Augmentation through Bayesian Model Selectioncited by 1
  8. 2026Regularized Meta-Learning for Improved Generalizationcited by 0
  9. 2026Variational predictive resamplingcited by 0