Indico has been upgraded to version 3.1. Details in the SSB
Mar 20 – 22, 2018
University of Washington Seattle
US/Pacific timezone

Uncertainty estimation of parameters with neural networks: Application to strong gravitational lensing

Not scheduled
Physics-Astronomy Auditorium A118 (University of Washington Seattle)

Physics-Astronomy Auditorium A118

University of Washington Seattle

Poster 6: Beyond the conventional tracking Poster


Dr Laurence Perreault Levasseur (Stanford University)


Recently we showed that deep learning can be used for model parameter estimation for strong gravitational lensing systems. Here we demonstrate a method for obtaining the uncertainties of these parameters. We use 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.

Primary author

Dr Laurence Perreault Levasseur (Stanford University)


Dr Yashar Hezaveh (Stanford University)

Presentation materials

There are no materials yet.