Pith. sign in

REVIEW 1 cited by

Uncertainties in Parameters Estimated with Neural Networks: Application to Strong Gravitational Lensing

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 1708.08843 v1 pith:5ONUKQ6Y submitted 2017-08-29 astro-ph.CO astro-ph.IM

Uncertainties in Parameters Estimated with Neural Networks: Application to Strong Gravitational Lensing

classification astro-ph.CO astro-ph.IM
keywords networksneuralparametersuncertaintieslensingparametercoverageestimation
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

In Hezaveh et al. 2017 we showed that deep learning can be used for model parameter estimation and trained convolutional neural networks to determine the parameters of strong gravitational lensing systems. Here we demonstrate a method for obtaining the uncertainties of these parameters. We review the framework of variational inference to obtain approximate posteriors of Bayesian neural networks and apply it to a network trained to estimate the parameters of the Singular Isothermal Ellipsoid plus external shear and total flux magnification. We show that the method can capture the uncertainties due to different levels of noise in the input data, as well as training and architecture-related errors made by the network. To evaluate the accuracy of the resulting uncertainties, we calculate the coverage probabilities of marginalized distributions for each lensing parameter. By tuning a single hyperparameter, the dropout rate, we obtain coverage probabilities approximately equal to the confidence levels for which they were calculated, resulting in accurate and precise uncertainty estimates. Our results suggest that neural networks can be a fast alternative to Monte Carlo Markov Chains for parameter uncertainty estimation in many practical applications, allowing more than seven orders of magnitude improvement in speed.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Line-of-sight shear in SLACS strong lenses II: validation tests with an extended sample

    astro-ph.CO 2025-12 conditional novelty 4.0

    Line-of-sight shear in 45 SLACS lenses is often larger than N-body expectations (mean |γ_LOS| ≈ 0.085–0.11), and adding mass octupoles or checking image properties does not remove the excess.