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

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.

arXiv commentsubmitted to ApJL

References in corpus (5)

• What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
• Fast Automated Analysis of Strong Gravitational Lenses with Convolutional Neural Networks
• Finding Strong Gravitational Lenses in the Kilo Degree Survey with Convolutional Neural Networks
• The SL2S Galaxy-scale Lens Sample. V. Dark Matter Halos and Stellar IMF of Massive Early-type Galaxies out to Redshift 0.8
• Finding strong lenses in CFHTLS using convolutional neural networks

Cited by in corpus (101)

• Deep Learning for Real-time Gravitational Wave Detection and Parameter Estimation: Results with Advanced LIGO Data
• Deep Neural Networks to Enable Real-time Multimessenger Astrophysics
• Galaxy Zoo DECaLS: Detailed Visual Morphology Measurements from Volunteers and Deep Learning for 314,000 Galaxies
• Cosmological constraints with deep learning from KiDS-450 weak lensing maps
• Non-parametric Star Formation History Reconstruction with Gaussian Processes I: Counting Major Episodes of Star Formation
• Surveying the reach and maturity of machine learning and artificial intelligence in astronomy
• Is every strong lens model unhappy in its own way? Uniform modelling of a sample of 13 quadruply+ imaged quasars
• The DAWES review 10: The impact of deep learning for the analysis of galaxy surveys
• Strong lensing time-delay cosmography in the 2020s
• Cosmological constraints from noisy convergence maps through deep learning
• Mining for Dark Matter Substructure: Inferring subhalo population properties from strong lenses with machine learning
• Cosmology From CMB Lensing and Delensed EE Power Spectra Using 2019-2020 SPT-3G Polarization Data
• A Systematic Comparison of Bayesian Deep Learning Robustness in Diabetic Retinopathy Tasks
• Simulation-Based Inference of Reionization Parameters From 3D Tomographic 21 cm Lightcone Images
• An Ensemble of Bayesian Neural Networks for Exoplanetary Atmospheric Retrieval
• Data-Driven Reconstruction of Gravitationally Lensed Galaxies using Recurrent Inference Machines
• HOLISMOKES -- I. Highly Optimised Lensing Investigations of Supernovae, Microlensing Objects, and Kinematics of Ellipticals and Spirals
• Photometric Redshift Estimation with a Convolutional Neural Network: NetZ
• Extensive Studies of the Neutron Star Equation of State from the Deep Learning Inference with the Observational Data Augmentation
• Detection and Parameter Estimation of Gravitational Waves from Binary Neutron-Star Mergers in Real LIGO Data using Deep Learning
• A Machine Learning Based Morphological Classification of 14,245 Radio AGNs Selected From The Best-Heckman Sample
• Parameters Estimation for the Cosmic Microwave Background with Bayesian Neural Networks
• The GCE in a New Light: Disentangling the $\gamma$-ray Sky with Bayesian Graph Convolutional Neural Networks
• Automated galaxy-galaxy strong lens modelling: no lens left behind
• Hierarchical Inference With Bayesian Neural Networks: An Application to Strong Gravitational Lensing
• The use of convolutional neural networks for modelling large optically-selected strong galaxy-lens samples
• Gravitational-Wave Detector Networks: Standard Sirens on Cosmology and Modified Gravity Theory
• Accurate Machine Learning Atmospheric Retrieval via a Neural Network Surrogate Model for Radiative Transfer
• Deep Learning the Morphology of Dark Matter Substructure
• From Images to Dark Matter: End-To-End Inference of Substructure From Hundreds of Strong Gravitational Lenses
• HOLISMOKES -- IV. Efficient mass modeling of strong lenses through deep learning
• Direct Detection of Dark Matter Substructure in Strong Lens Images with Convolutional Neural Networks
• Differentiable Strong Lensing: Uniting Gravity and Neural Nets through Differentiable Probabilistic Programming
• Translating neutron star observations to nuclear symmetry energy via artificial neural networks
• Breaking baryon-cosmology degeneracy with the electron density power spectrum
• Large-Scale Gravitational Lens Modeling with Bayesian Neural Networks for Accurate and Precise Inference of the Hubble Constant
• TDCOSMO X. Automated Modeling of 9 Strongly Lensed Quasars and Comparison Between Lens Modeling Software
• Designing Accurate Emulators for Scientific Processes using Calibration-Driven Deep Models
• Simulation-based inference of dynamical galaxy cluster masses with 3D convolutional neural networks
• Estimating the warm dark matter mass from strong lensing images with truncated marginal neural ratio estimation
• Strong lens modelling: comparing and combining Bayesian neural networks and parametric profile fitting
• Deep Horizon; a machine learning network that recovers accreting black hole parameters
• Extracting the Subhalo Mass Function from Strong Lens Images with Image Segmentation
• Robust Simulation-Based Inference in Cosmology with Bayesian Neural Networks
• GAlaxy Light profile convolutional neural NETworks (GaLNets). I. fast and accurate structural parameters for billion galaxy samples
• A machine learning based approach to gravitational lens identification with the International LOFAR Telescope
• Variational Inference as an alternative to MCMC for parameter estimation and model selection
• SLITronomy: towards a fully wavelet-based strong lensing inversion technique
• Cleaning our own Dust: Simulating and Separating Galactic Dust Foregrounds with Neural Networks
• MillimeterDL: Deep Learning Simulations of the Microwave Sky
• Photometric redshift estimates using Bayesian neural networks in the CSST survey
• Constraining the Reionization History using Bayesian Normalizing Flows
• Auto-identification of unphysical source reconstructions in strong gravitational lens modelling
• HOLISMOKES -- IX. Neural network inference of strong-lens parameters and uncertainties from ground-based images
• HOLISMOKES -- X. Comparison between neural network and semi-automated traditional modeling of strong lenses
• Decoding Dark Matter Substructure without Supervision
• Benchmarking Bayesian Deep Learning on Diabetic Retinopathy Detection Tasks
• LeMoN: Lens Modelling with Neural networks -- I. Automated modelling of strong gravitational lenses with Bayesian Neural Networks
• A Framework for Obtaining Accurate Posteriors of Strong Gravitational Lensing Parameters with Flexible Priors and Implicit Likelihoods using Density Estimation
• GaMPEN: A Machine Learning Framework for Estimating Bayesian Posteriors of Galaxy Morphological Parameters
• Reconstructing Cosmic Polarization Rotation with ResUNet-CMB
• Deep learning the astrometric signature of dark matter substructure
• AI-driven spatio-temporal engine for finding gravitationally lensed type Ia supernovae
• Pixelated Reconstruction of Foreground Density and Background Surface Brightness in Gravitational Lensing Systems using Recurrent Inference Machines
• Modeling lens potentials with continuous neural fields in galaxy-scale strong lenses
• Deep Learning in Wide-field Surveys: Fast Analysis of Strong Lenses in Ground-based Cosmic Experiments
• Galaxy Spectra neural Network (GaSNet). II. Using Deep Learning for Spectral Classification and Redshift Predictions
• Using Bayesian Deep Learning to infer Planet Mass from Gaps in Protoplanetary Disks
• Forklens: Accurate weak-lensing shear measurement with deep learning
• Parameter estimation of microlensed gravitational waves with Conditional Variational Autoencoders
• Inferring subhalo effective density slopes from strong lensing observations with neural likelihood-ratio estimation
• Inferring stellar parameters and their uncertainties from high-resolution spectroscopy using invertible neural networks
• Deep Learning Blazar Classification based on Multi-frequency Spectral Energy Distribution Data
• Estimating Photometric Redshifts for Galaxies from the DESI Legacy Imaging Surveys with Bayesian Neural Networks Trained by DESI EDR
• Inferring redshift and galaxy properties via a multi-task neural net with probabilistic outputs: An application to simulated MOONS spectra
• Anisotropic strong lensing as a probe of dark matter self-interactions
• Peculiar Velocity Estimation from Kinetic SZ Effect using Deep Neural Networks
• AGNet: Weighing Black Holes with Deep Learning
• Accurately Estimating Redshifts from CSST Slitless Spectroscopic Survey using Deep Learning
• Lens Modeling of STRIDES Strongly Lensed Quasars using Neural Posterior Estimation
• Heteroscedastic Calibration of Uncertainty Estimators in Deep Learning
• Dark Matter Subhalos, Strong Lensing and Machine Learning
• Constraining Cosmology with Double-source-plane Strong Gravitational Lenses from the AGEL Survey
• A proof-of-concept neural network for inferring parameters of a black hole from partial interferometric images of its shadow
• Meaningful uncertainties from deep neural network surrogates of large-scale numerical simulations
• Measuring the Substructure Mass Power Spectrum of 23 SLACS Strong Galaxy-Galaxy Lenses with Convolutional Neural Networks
• Prediction intervals for Deep Neural Networks
• A Modular Deep Learning Pipeline for Galaxy-Scale Strong Gravitational Lens Detection and Modeling
• Bayesian and Convolutional Networks for Hierarchical Morphological Classification of Galaxies
• Accelerating lensed quasar discovery and modeling with physics-informed variational autoencoders
• A Bayesian neural network predicts the dissolution of compact planetary systems
• kSZ Pairwise Velocity Reconstruction with Machine Learning
• Reliable Uncertainties for Bayesian Neural Networks using Alpha-divergences
• Understanding the predication mechanism of deep learning through error propagation among parameters in strong lensing case
• AGNet: Weighing Black Holes with Machine Learning
• Parameters Estimation from the 21 cm signal using Variational Inference
• Deep inference of simulated strong lenses in ground-based surveys
• Genetic-algorithm-optimized neural networks for gravitational wave classification
• A New Method for the Reconstruction of Strongly Lensed Galaxies with Resolved Kinematics
• Enhancing Gravitational Lens Study with Deep Learning: A Study on Effects of Dropout Regularization
• Machine Learning for Nanohertz Gravitational Wave Detection and Parameter Estimation with Pulsar Timing Array