Epistemic Uncertainty and Observation Noise with the Neural Tangent Kernel

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Calvo-Ordoñez, Sergio, Palla, Konstantina, Ciosek, Kamil
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914945773338624
author Calvo-Ordoñez, Sergio
Palla, Konstantina
Ciosek, Kamil
author_facet Calvo-Ordoñez, Sergio
Palla, Konstantina
Ciosek, Kamil
contents Recent work has shown that training wide neural networks with gradient descent is formally equivalent to computing the mean of the posterior distribution in a Gaussian Process (GP) with the Neural Tangent Kernel (NTK) as the prior covariance and zero aleatoric noise \parencite{jacot2018neural}. In this paper, we extend this framework in two ways. First, we show how to deal with non-zero aleatoric noise. Second, we derive an estimator for the posterior covariance, giving us a handle on epistemic uncertainty. Our proposed approach integrates seamlessly with standard training pipelines, as it involves training a small number of additional predictors using gradient descent on a mean squared error loss. We demonstrate the proof-of-concept of our method through empirical evaluation on synthetic regression.
format Preprint
id arxiv_https___arxiv_org_abs_2409_03953
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Epistemic Uncertainty and Observation Noise with the Neural Tangent Kernel
Calvo-Ordoñez, Sergio
Palla, Konstantina
Ciosek, Kamil
Machine Learning
Recent work has shown that training wide neural networks with gradient descent is formally equivalent to computing the mean of the posterior distribution in a Gaussian Process (GP) with the Neural Tangent Kernel (NTK) as the prior covariance and zero aleatoric noise \parencite{jacot2018neural}. In this paper, we extend this framework in two ways. First, we show how to deal with non-zero aleatoric noise. Second, we derive an estimator for the posterior covariance, giving us a handle on epistemic uncertainty. Our proposed approach integrates seamlessly with standard training pipelines, as it involves training a small number of additional predictors using gradient descent on a mean squared error loss. We demonstrate the proof-of-concept of our method through empirical evaluation on synthetic regression.
title Epistemic Uncertainty and Observation Noise with the Neural Tangent Kernel
topic Machine Learning
url https://arxiv.org/abs/2409.03953