Input Uncertainty

Input uncertainty, encompassing inaccuracies or randomness in model inputs, is a critical challenge in machine learning, hindering reliable predictions and robust decision-making. Current research focuses on developing methods to quantify and propagate input uncertainty through various model architectures, including neural networks and Gaussian processes, often employing techniques like conformal prediction, bootstrapping, and mixture-of-experts models to achieve better calibration and efficiency. Addressing input uncertainty is crucial for improving the reliability and safety of machine learning applications across diverse fields, from robotics and finance to medical imaging and language processing, where input noise or ambiguity is inherent.

Papers

November 13, 2024

DeepUQ: Assessing the Aleatoric Uncertainties from two Deep Learning Methods
Rebecca Nevin, Aleksandra Ćiprijanović, Brian D. Nord
Deep Learning NCD Method Aleatoric Uncertainty Input Uncertainty

September 30, 2024

End-to-End Conformal Calibration for Optimization Under Uncertainty
Christopher Yeh, Nicolas Christianson, Alan Wu, Adam Wierman, Yisong Yue
Optimization Purpose High Uncertainty Anticipation Robustness Guarantee Conformal Regression Calibrated Uncertainty Input Uncertainty Input Convex Neural Network

August 7, 2024

Question Rephrasing for Quantifying Uncertainty in Large Language Models: Applications in Molecular Chemistry Tasks
Zizhang Chen, Pengyu Hong, Sandeep Madireddy
Large Language Model Financial Application Uncertainty Quantification Uncertainty Estimation Input Uncertainty Question Mark Chemical Named Entity Recognition

June 26, 2024

Unified Uncertainties: Combining Input, Data and Model Uncertainty into a Single Formulation
Matias Valdenegro-Toro, Ivo Pascal de Jong, Marco Zullich
Raw Data High Uncertainty Anticipation Model Uncertainty Decision Boundary Rich Input Prediction Uncertainty New Formulation Input Uncertainty Unified Uncertainty

June 11, 2024

Conformalized Teleoperation: Confidently Mapping Human Inputs to High-Dimensional Robot Actions
Michelle Zhao, Reid Simmons, Henny Admoni, Andrea Bajcsy
High Dimensional Teleoperation Mode Input Uncertainty Assistive Robotic Arm Assistive Teleoperation High Dimensional Quantile Regression

April 29, 2024

Orthogonal Bootstrap: Efficient Simulation of Input Uncertainty
Kaizhao Liu, Jose Blanchet, Lexing Ying, Yiping Lu
Fast Simulation Input Uncertainty Orthogonal Statistical Learning Model Replication Joint Constraint

February 13, 2024

Uncertainty Quantification via Stable Distribution Propagation
Felix Petersen, Aashwin Mishra, Hilde Kuehne, Christian Borgelt, Oliver Deussen, Mikhail Yurochkin
Neural Network Uncertainty Quantification RSD Difference of Gaussian Input Uncertainty Local Linearity Alpha Stable

December 13, 2023

Training of Neural Networks with Uncertain Data: A Mixture of Experts Approach
Lucas Luttner
Neural Network Training Data Mixture Component Expert Knowledge Aleatoric Uncertainty Data Uncertainty Input Uncertainty

October 31, 2023

Efficient Robust Bayesian Optimization for Arbitrary Uncertain Inputs
Lin Yang, Junlong Lyu, Wenlong Lyu, Zhitang Chen
Bayesian Optimization Sample Efficient Input Uncertainty Stochastic Robust Optimization Robust Bayesian Optimization

September 28, 2023

Propagation and Attribution of Uncertainty in Medical Imaging Pipelines
Leonhard F. Feiner, Martin J. Menten, Kerstin Hammernik, Paul Hager, Wenqi Huang, Daniel Rueckert, Rickmer F. Braren, Georgios Kaissis
High Uncertainty Anticipation Uncertainty Estimation Source Attribution Fast Propagation Input Uncertainty Imaging Pipeline

September 12, 2023

RT-LM: Uncertainty-Aware Resource Management for Real-Time Inference of Language Models
Yufei Li, Zexin Li, Wei Yang, Cong Liu
Language Model Inference Latency Real Time Inference Input Uncertainty

May 19, 2023

Bayesian approach to Gaussian process regression with uncertain inputs
Dongwei Ye, Mengwu Guo
Gaussian Process Bayesian Inference Gaussian Process Regression Bayesian Approach Input Uncertainty

May 16, 2023

Sorting and Hypergraph Orientation under Uncertainty with Predictions
Thomas Erlebach, Murilo Santos de Lima, Nicole Megow, Jens Schlöter
Human Prediction High Uncertainty Anticipation Worst Case Auxiliary Sorting Network Rearranges Learning Augmented Algorithm Input Uncertainty Veracity Prediction Uncertainty Aware Exploration

January 27, 2023

Bounding Box-based Multi-objective Bayesian Optimization of Risk Measures under Input Uncertainty
Yu Inatsu, Shion Takeno, Hiroyuki Hanada, Kazuki Iwata, Ichiro Takeuchi
Bounding Box Pareto Optimal Pareto Front Black Box Function Multi Objective Bayesian Optimization Risk Measure Input Uncertainty Bayes Risk

October 18, 2022

Disentangling the Predictive Variance of Deep Ensembles through the Neural Tangent Kernel
Seijin Kobayashi, Pau Vilimelis Aceituno, Johannes von Oswald
Inductive Bias Neural Tangent Kernel Deep Ensemble Input Uncertainty

July 21, 2022

Generating and Detecting True Ambiguity: A Forgotten Danger in DNN Supervision Testing
Michael Weiss, André García Gómez, Paolo Tonella
Adversarial Input Ambiguous Input Input Uncertainty Hidden Risk

February 25, 2022

A Robust Multi-Objective Bayesian Optimization Framework Considering Input Uncertainty
J. Qing, I. Couckuyt, T. Dhaene
Bayesian Optimization Multi Objective Optimization Multi Objective Bayesian Optimization Input Uncertainty

December 5, 2021

Diverse, Global and Amortised Counterfactual Explanations for Uncertainty Estimates
Dan Ley, Umang Bhatt, Adrian Weller
Uncertainty Estimation Counterfactual Explanation World Event Diverse Set Superficial Clue Input Uncertainty

Input Uncertainty

Papers

DeepUQ: Assessing the Aleatoric Uncertainties from two Deep Learning Methods

End-to-End Conformal Calibration for Optimization Under Uncertainty

Question Rephrasing for Quantifying Uncertainty in Large Language Models: Applications in Molecular Chemistry Tasks

Unified Uncertainties: Combining Input, Data and Model Uncertainty into a Single Formulation

Conformalized Teleoperation: Confidently Mapping Human Inputs to High-Dimensional Robot Actions

Orthogonal Bootstrap: Efficient Simulation of Input Uncertainty

Uncertainty Quantification via Stable Distribution Propagation

Training of Neural Networks with Uncertain Data: A Mixture of Experts Approach

Efficient Robust Bayesian Optimization for Arbitrary Uncertain Inputs

Propagation and Attribution of Uncertainty in Medical Imaging Pipelines

RT-LM: Uncertainty-Aware Resource Management for Real-Time Inference of Language Models

Bayesian approach to Gaussian process regression with uncertain inputs

Sorting and Hypergraph Orientation under Uncertainty with Predictions

Bounding Box-based Multi-objective Bayesian Optimization of Risk Measures under Input Uncertainty

Disentangling the Predictive Variance of Deep Ensembles through the Neural Tangent Kernel

Generating and Detecting True Ambiguity: A Forgotten Danger in DNN Supervision Testing

A Robust Multi-Objective Bayesian Optimization Framework Considering Input Uncertainty

Diverse, Global and Amortised Counterfactual Explanations for Uncertainty Estimates