Aleatoric and Epistemic Uncertainty in Machine Learning: An Introduction to Concepts and Methods

Abstract

The notion of uncertainty is of major importance in machine learning and constitutes a key element of machine learning methodology. In line with the statistical tradition, uncertainty has long been perceived as almost synonymous with standard probability and probabilistic predictions. Yet, due to the steadily increasing relevance of machine learning for practical applications and related issues such as safety requirements, new problems and challenges have recently been identified by machine learning scholars, and these problems may call for new methodological developments. In particular, this includes the importance of distinguishing between (at least) two different types of uncertainty, often referred to as aleatoric and epistemic. In this paper, we provide an introduction to the topic of uncertainty in machine learning as well as an overview of attempts so far at handling uncertainty in general and formalizing this distinction in particular.

59 pages

References in corpus (2)

• Unifying Practical Uncertainty Representations: I. Generalized P-Boxes
• Consistent Algorithms for Multiclass Classification with a Reject Option

Cited by in corpus (117)

• Remote patient monitoring using artificial intelligence: Current state, applications, and challenges
• Uncertainty Quantification in Scientific Machine Learning: Methods, Metrics, and Comparisons
• Uncertainty Quantification in Machine Learning for Engineering Design and Health Prognostics: A Tutorial
• Classifier Calibration: A survey on how to assess and improve predicted class probabilities
• A review of predictive uncertainty estimation with machine learning
• Current and future directions in network biology
• A Survey on Epistemic (Model) Uncertainty in Supervised Learning: Recent Advances and Applications
• Lymphoma segmentation from 3D PET-CT images using a deep evidential network
• Ensemble Quantile Networks: Uncertainty-Aware Reinforcement Learning with Applications in Autonomous Driving
• Finite electro-elasticity with physics-augmented neural networks
• How to Evaluate Uncertainty Estimates in Machine Learning for Regression?
• Knowledge graph enhanced retrieval-augmented generation for failure mode and effects analysis
• Look Before You Leap: An Exploratory Study of Uncertainty Measurement for Large Language Models
• Application of belief functions to medical image segmentation: A review
• Artificial Neural Networks Modelling of Wall Pressure Spectra Beneath Turbulent Boundary Layers
• Assessments of epistemic uncertainty using Gaussian stochastic weight averaging for fluid-flow regression
• Uncertainty-aware Surrogate Models for Airfoil Flow Simulations with Denoising Diffusion Probabilistic Models
• Bayesian Physics Informed Neural Networks for Data Assimilation and Spatio-Temporal Modelling of Wildfires
• Reliability and Interpretability in Science and Deep Learning
• Diffusion Variational Autoencoder for Tackling Stochasticity in Multi-Step Regression Stock Price Prediction
• Explainability through uncertainty: Trustworthy decision-making with neural networks
• Learning Uncertainty with Artificial Neural Networks for Improved Predictive Process Monitoring
• Learning Curves for Decision Making in Supervised Machine Learning: A Survey
• Uncertainty Aware Training to Improve Deep Learning Model Calibration for Classification of Cardiac MR Images
• The Unreasonable Effectiveness of Deep Evidential Regression
• M3BAT: Unsupervised Domain Adaptation for Multimodal Mobile Sensing with Multi-Branch Adversarial Training
• Gradient-descent quantum process tomography by learning Kraus operators
• Uncertainty Quantification and Propagation in Surrogate-based Bayesian Inference
• Formulation Graphs for Mapping Structure-Composition of Battery Electrolytes to Device Performance
• Bridging the gap: Towards an Expanded Toolkit for AI-driven Decision-Making in the Public Sector
• Low-Order Flow Reconstruction and Uncertainty Quantification in Disturbed Aerodynamics Using Sparse Pressure Measurements
• DC-Check: A Data-Centric AI checklist to guide the development of reliable machine learning systems
• Audit to Forget: A Unified Method to Revoke Patients' Private Data in Intelligent Healthcare
• Calibration in Machine Learning Uncertainty Quantification: beyond consistency to target adaptivity
• Some models are useful, but how do we know which ones? Towards a unified Bayesian model taxonomy
• Distributionally-Informed Recommender System Evaluation
• Quantifying Uncertainty in Deep Learning Approaches to Radio Galaxy Classification
• Artificial Intelligence for Science in Quantum, Atomistic, and Continuum Systems
• Addressing contingency in algorithmic (mis)information classification: Toward a responsible machine learning agenda
• On the effects of biased quantum random numbers on the initialization of artificial neural networks
• Ringdown mode amplitudes of precessing binary black holes
• How False Data Affects Machine Learning Models in Electrochemistry?
• Fairness through Aleatoric Uncertainty
• Empowering Bayesian Neural Networks with Functional Priors through Anchored Ensembling for Mechanics Surrogate Modeling Applications
• A Hybrid Probabilistic Battery Health Management Approach for Robust Inspection Drone Operations
• Sample-efficient estimation of entanglement entropy through supervised learning
• Accounting for Model Uncertainty in Algorithmic Discrimination
• Improved conformalized quantile regression
• Attention-based Multi-fidelity Machine Learning Model for Computational Fractional Flow Reserve Assessment
• Loss-Aversively Fair Classification
• Attenuation Coefficient Estimation for PET/MRI With Bayesian Deep Learning pseudo-CT and Maximum Likelihood Estimation of Activity and Attenuation
• A deep learning model for chemical shieldings in molecular organic solids including anisotropy
• Using Bayesian Deep Learning to infer Planet Mass from Gaps in Protoplanetary Disks
• A Lightweight Method for Modeling Confidence in Recommendations with Learned Beta Distributions
• Integrating Uncertainty into Neural Network-based Speech Enhancement
• Uncertainty Quantification of Wind Gust Predictions in the Northeast United States: An Evidential Neural Network and Explainable Artificial Intelligence Approach
• Detection of Dipole Modulation in CMB Temperature Anisotropy Maps from WMAP and Planck using Artificial Intelligence
• A Temporal Stochastic Bias Correction using a Machine Learning Attention model
• Uncertainty-Aware Self-supervised Neural Network for Liver $T_{1\rho }$ Mapping with Relaxation Constraint
• Performance of a GPU- and Time-Efficient Pseudo 3D Network for Magnetic Resonance Image Super-Resolution and Motion Artifact Reduction
• Inference of the optical depth to reionization $\tau$ from $\text{Planck}$ CMB maps with convolutional neural networks
• Deep Huber quantile regression networks
• Euclid preparation. LXVIII. Extracting physical parameters from galaxies with machine learning
• Aleatoric uncertainty for Errors-in-Variables models in deep regression
• Efficient ensemble uncertainty estimation in Gaussian Processes Regression
• A Probabilistic Model for Aircraft in Climb using Monotonic Functional Gaussian Process Emulators
• Categorising the World into Local Climate Zones -- Towards Quantifying Labelling Uncertainty for Machine Learning Models
• Fast and reliable uncertainty quantification with neural network ensembles for industrial image classification
• Ensemble Neural Networks for Remaining Useful Life (RUL) Prediction
• Accurately Estimating Redshifts from CSST Slitless Spectroscopic Survey using Deep Learning
• Distributional Actor-Critic Ensemble for Uncertainty-Aware Continuous Control
• Formalising Anti-Discrimination Law in Automated Decision Systems
• A Survey on Event Prediction Methods from a Systems Perspective: Bringing Together Disparate Research Areas
• Graph-accelerated non-intrusive polynomial chaos expansion using partially tensor-structured quadrature rules for uncertainty quantification
• A Priori Uncertainty Quantification of Reacting Turbulence Closure Models using Bayesian Neural Networks
• Abstraction Alignment: Comparing Model-Learned and Human-Encoded Conceptual Relationships
• Shell quenching in nuclear charge radii based on Monte Carlo dropout Bayesian neural network
• Cross-model Fairness: Empirical Study of Fairness and Ethics Under Model Multiplicity
• Combined Data and Deep Learning Model Uncertainties: An Application to the Measurement of Solid Fuel Regression Rate
• Interpretable machine learning approach for electron antineutrino selection in a large liquid scintillator detector
• Fairness in Ranking under Disparate Uncertainty
• Risks of ignoring uncertainty propagation in AI-augmented security pipelines
• Bayesian Neural Networks for Virtual Flow Metering: An Empirical Study
• A Mapping of Assurance Techniques for Learning Enabled Autonomous Systems to the Systems Engineering Lifecycle
• Reliable Classification with Conformal Learning and Interval-Type 2 Fuzzy Sets
• EuroPED-NN: Uncertainty aware surrogate model
• Evaluating the Quality of the Quantified Uncertainty for (Re)Calibration of Data-Driven Regression Models
• A gradient-enhanced univariate dimension reduction method for uncertainty propagation
• Disentangling Aleatoric and Epistemic Uncertainty in Physics-Informed Neural Networks. Application to Insulation Material Degradation Prognostics
• From ANN to BNN: Inferring Reionization Parameters using Uncertainty-aware Emulators of 21-cm Summaries
• Uncertainty-Aware Knowledge Distillation for Compact and Efficient 6DoF Pose Estimation
• Uncertainties of Satellite-based Essential Climate Variables from Deep Learning
• Interactive incremental learning of generalizable skills with local trajectory modulation
• Selecting fitted models under epistemic uncertainty using a stochastic process on quantile functions
• Confident Naturalness Explanation (CNE): A Framework to Explain and Assess Patterns Forming Naturalness
• Beyond Accuracy: An Empirical Study of Uncertainty Estimation in Imputation
• A Frank System for Co-Evolutionary Hybrid Decision-Making
• Robust Partial-Label Learning by Leveraging Class Activation Values
• An Efficient Model-Agnostic Approach for Uncertainty Estimation in Data-Restricted Pedometric Applications
• Uncertainty quantification and stability of neural operators for prediction of three-dimensional turbulence
• The Impact of Machine Learning Uncertainty on the Robustness of Counterfactual Explanations
• Advancing Positron Emission Tomography Image Quantification: Artificial Intelligence-Driven Methods, Clinical Challenges, and Emerging Opportunities in Long-Axial Field-of-View Positron Emission Tomography/Computed Tomography Imaging
• Bayesian-EUCLID: discovering hyperelastic material laws with uncertainties
• Assimilation of machine learning-predicted nitrate to improve the quality of phytoplankton forecasting in the shelf sea environment
• Multivariate Uncertainty Quantification with Tomographic Quantile Forests
• Autonomous Uncertainty Quantification for Computational Point-of-care Sensors
• The SLAM Confidence Trap
• TabPFN Extensions for Interpretable Geotechnical Modelling
• Navigating Uncertainties in Machine Learning for Structural Dynamics: A Comprehensive Survey of Probabilistic and Non-Probabilistic Approaches in Forward and Inverse Problems
• CUQ-GNN: Committee-based Graph Uncertainty Quantification using Posterior Networks
• ILeSiA: Interactive Learning of Robot Situational Awareness from Camera Input
• Can Uncertainty Quantification Improve Learned Index Benefit Estimation?
• A context-adaptive policy framework for robust and reactive robotic manipulation via uncertainty-aware imitation learning
• Synergistic Fusion of Multi-Source Knowledge via Evidence Theory for High-Entropy Alloy Discovery
• Conceptualizing Uncertainty: A Concept-based Approach to Explaining Uncertainty
• Detecting underdetermination in parameterized quantum circuits
• An Analysis of Temporal Dropout in Earth Observation Time Series for Regression Tasks