Epistemic Uncertainty

Epistemic uncertainty, representing a model's lack of knowledge rather than inherent data noise (aleatoric uncertainty), is a crucial area of research in machine learning, aiming to improve the reliability and trustworthiness of AI systems. Current efforts focus on developing methods to quantify and calibrate epistemic uncertainty within various model architectures, including Bayesian neural networks, deep ensembles, and energy-based models, often applied to tasks like out-of-distribution detection and safe reinforcement learning. Accurate epistemic uncertainty quantification is vital for building robust and reliable AI systems across diverse applications, from medical diagnosis to autonomous driving, by enabling models to identify their limitations and avoid overconfident predictions in uncertain situations. The field is actively addressing challenges such as disentangling aleatoric and epistemic uncertainty and mitigating the "epistemic uncertainty collapse" observed in large models.

Papers

March 23, 2023

Laplacian Segmentation Networks: Improved Epistemic Uncertainty from Spatial Aleatoric Uncertainty
Kilian Zepf, Selma Wanna, Marco Miani, Juston Moore, Jes Frellsen, Søren Hauberg, Aasa Feragen, Frederik Warburg
Image Segmentation Epistemic Uncertainty Aleatoric Uncertainty Laplacian Representation Inaccurate Segmentation

March 10, 2023

Decision-Making Under Uncertainty: Beyond Probabilities
Thom Badings, Thiago D. Simão, Marnix Suilen, Nils Jansen
Reinforcement Learning Markov Decision Process High Uncertainty Anticipation Epistemic Uncertainty Aleatoric Uncertainty Partial Observability Relative Probability

March 7, 2023

Exploration via Epistemic Value Estimation
Simon Schmitt, John Shawe-Taylor, Hado van Hasselt
Reinforcement Learning Environment Exploration Epistemic Uncertainty Efficient Exploration

March 4, 2023

Wasserstein Actor-Critic: Directed Exploration via Optimism for Continuous-Actions Control
Amarildo Likmeta, Matteo Sacco, Alberto Maria Metelli, Marcello Restelli
Reinforcement Learning Q Learning Epistemic Uncertainty Distributional Reinforcement Learning Wasserstein Dictionary Directed Exploration Continuous Action

February 28, 2023

Playing to Learn, or to Keep Secret: Alternating-Time Logic Meets Information Theory
Masoud Tabatabaei, Wojciech Jamroga
Multi Agent System Epistemic Uncertainty Intelligent Agent Information Theory Epistemic Space Alternating Time Temporal Logic

February 26, 2023

Navigating the Grey Area: How Expressions of Uncertainty and Overconfidence Affect Language Models
Kaitlyn Zhou, Dan Jurafsky, Tatsunori Hashimoto
Language Model High Uncertainty Anticipation Epistemic Uncertainty Uncertain Reasoning Human Expression Epistemic State Evidential Reasoning Relative Stance Bias

February 17, 2023

Learning to Forecast Aleatoric and Epistemic Uncertainties over Long Horizon Trajectories
Aastha Acharya, Rebecca Russell, Nisar R. Ahmed
LeArning Abstract Autonomous Agent World Model Epistemic Uncertainty Agent Trajectory Non Gaussian Aleatoric Uncertainty Long Horizon Trajectory

February 16, 2023

A Review of Uncertainty Estimation and its Application in Medical Imaging
Ke Zou, Zhihao Chen, Xuedong Yuan, Xiaojing Shen, Meng Wang, Huazhu Fu
Application Proficiency Uncertainty Estimation Medical Imaging Epistemic Uncertainty Aleatoric Uncertainty

February 6, 2023

Dropout Injection at Test Time for Post Hoc Uncertainty Quantification in Neural Networks
Emanuele Ledda, Giorgio Fumera, Fabio Roli
Neural Network Epistemic Uncertainty Structured Dropout Test Time Dropout Layer Invariant Dropout

February 2, 2023

Normalizing Flow Ensembles for Rich Aleatoric and Epistemic Uncertainty Modeling
Lucas Berry, David Meger
Normalizing Flow Epistemic Uncertainty Aleatoric Uncertainty Structured Distribution Non Gaussian Aleatoric Uncertainty

January 11, 2023

How Does Traffic Environment Quantitatively Affect the Autonomous Driving Prediction?
Wenbo Shao, Yanchao Xu, Jun Li, Chen Lv, Weida Wang, Hong Wang
Autonomous Driving Trajectory Prediction Epistemic Uncertainty Traffic Environment

December 1, 2022

An introduction to optimization under uncertainty -- A short survey
Keivan Shariatmadar, Kaizheng Wang, Calvin R. Hubbard, Hans Hallez, David Moens
Optimization Purpose High Uncertainty Anticipation Epistemic Uncertainty Gentle Introduction Robust Solution

November 30, 2022

One Risk to Rule Them All: A Risk-Sensitive Perspective on Model-Based Offline Reinforcement Learning
Marc Rigter, Bruno Lacerda, Nick Hawes
Offline Reinforcement Learning Epistemic Uncertainty Risk Sensitive High Quality Risk Description Model Based Offline Reinforcement Learning Risk Sensitive Reinforcement Learning High Epistemic Uncertainty

November 11, 2022

Disentangled Uncertainty and Out of Distribution Detection in Medical Generative Models
Kumud Lakara, Matias Valdenegro-Toro
Uncertainty Quantification Distribution Detection Image to Image Translation Epistemic Uncertainty High Uncertainty Value Medical Generative Uncertainty Disentanglement

October 21, 2022

E-MCTS: Deep Exploration in Model-Based Reinforcement Learning by Planning with Epistemic Uncertainty
Yaniv Oren, Matthijs T. J. Spaan, Wendelin Böhmer
Model Based Reinforcement Learning Monte Carlo Tree Search Epistemic Uncertainty

October 20, 2022

Uncertainty Disentanglement with Non-stationary Heteroscedastic Gaussian Processes for Active Learning
Zeel B Patel, Nipun Batra, Kevin Murphy
Active Learning Gaussian Process Epistemic Uncertainty Bayesian Nonparametric Uncertainty Disentanglement

October 19, 2022

Propagating Variational Model Uncertainty for Bioacoustic Call Label Smoothing
Georgios Rizos, Jenna Lawson, Simon Mitchell, Pranay Shah, Xin Wen, Cristina Banks-Leite, Robert Ewers, Bjoern W. Schuller
Bayesian Neural Network Epistemic Uncertainty Predictive Uncertainty Uncertainty Aware Uncertainty Propagation

October 18, 2022

An out-of-distribution discriminator based on Bayesian neural network epistemic uncertainty
Ethan Ancell, Christopher Bennett, Bert Debusschere, Sapan Agarwal, Park Hays, T. Patrick Xiao
Neural Network Generative Adversarial Network Bayesian Neural Network Epistemic Uncertainty

October 12, 2022

Epistemic Uncertainty

Papers

Laplacian Segmentation Networks: Improved Epistemic Uncertainty from Spatial Aleatoric Uncertainty

Decision-Making Under Uncertainty: Beyond Probabilities

Exploration via Epistemic Value Estimation

Wasserstein Actor-Critic: Directed Exploration via Optimism for Continuous-Actions Control

Playing to Learn, or to Keep Secret: Alternating-Time Logic Meets Information Theory

Navigating the Grey Area: How Expressions of Uncertainty and Overconfidence Affect Language Models

Learning to Forecast Aleatoric and Epistemic Uncertainties over Long Horizon Trajectories

A Review of Uncertainty Estimation and its Application in Medical Imaging

Dropout Injection at Test Time for Post Hoc Uncertainty Quantification in Neural Networks

Normalizing Flow Ensembles for Rich Aleatoric and Epistemic Uncertainty Modeling

How Does Traffic Environment Quantitatively Affect the Autonomous Driving Prediction?

An introduction to optimization under uncertainty -- A short survey

One Risk to Rule Them All: A Risk-Sensitive Perspective on Model-Based Offline Reinforcement Learning

Disentangled Uncertainty and Out of Distribution Detection in Medical Generative Models

E-MCTS: Deep Exploration in Model-Based Reinforcement Learning by Planning with Epistemic Uncertainty

Uncertainty Disentanglement with Non-stationary Heteroscedastic Gaussian Processes for Active Learning

Propagating Variational Model Uncertainty for Bioacoustic Call Label Smoothing

An out-of-distribution discriminator based on Bayesian neural network epistemic uncertainty

Quantifying Uncertainty with Probabilistic Machine Learning Modeling in Wireless Sensing

Probabilities Are Not Enough: Formal Controller Synthesis for Stochastic Dynamical Models with Epistemic Uncertainty