Deep inference of simulated strong lenses in ground-based surveys
arXiv:2501.08524 · doi:10.1088/1475-7516/2025/05/053
Abstract
The large number of strong lenses discoverable in future astronomical surveys will likely enhance the value of strong gravitational lensing as a cosmic probe of dark energy and dark matter. However, leveraging the increased statistical power of such large samples will require further development of automated lens modeling techniques. We show that deep learning and simulation-based inference (SBI) methods produce informative and reliable estimates of parameter posteriors for strong lensing systems in ground-based surveys. We present the examination and comparison of two approaches to lens parameter estimation for strong galaxy-galaxy lenses -- Neural Posterior Estimation (NPE) and Bayesian Neural Networks (BNNs). We perform inference on 1-, 5-, and 12-parameter lens models for ground-based imaging data that mimics the Dark Energy Survey (DES). We find that NPE outperforms BNNs, producing posterior distributions that are more accurate, precise, and well-calibrated for most parameters. For the 12-parameter NPE model, the calibration is consistently within 10\% of optimal calibration for all parameters, while the BNN is rarely within 20\% of optimal calibration for any of the parameters. Similarly, residuals for most of the parameters are smaller (by up to an order of magnitude) with the NPE model than the BNN model. This work takes important steps in the systematic comparison of methods for different levels of model complexity.
Updated version to match published version
References in corpus (94)
- Adam: A Method for Stochastic Optimization
- ADADELTA: An Adaptive Learning Rate Method
- LSST: from Science Drivers to Reference Design and Anticipated Data Products
- On Calibration of Modern Neural Networks
- Variational Inference with Normalizing Flows
- Visualization in Bayesian workflow
- The Dark Energy Survey: more than dark energy - an overview
- Hands-on Bayesian Neural Networks -- a Tutorial for Deep Learning Users
- The Sloan Lens ACS Survey. X. Stellar, Dynamical, and Total Mass Correlations of Massive Early-type Galaxies
- The Dark Energy Survey Data Release 1
- Strong Lensing by Galaxies
- The Sloan Lens ACS Survey. V. The Full ACS Strong-Lens Sample
- Dissecting the Gravitational Lens B1608+656. II. Precision Measurements of the Hubble Constant, Spatial Curvature, and the Dark Energy Equation of State
- Gravitationally lensed quasars and supernovae in future wide-field optical imaging surveys
- A Comparative Review of Dimension Reduction Methods in Approximate Bayesian Computation
- Two accurate time-delay distances from strong lensing: Implications for cosmology
- The population of galaxy-galaxy strong lenses in forthcoming optical imaging surveys
- The Dark Energy Survey Data Release 2
- Lenstronomy: multi-purpose gravitational lens modelling software package
- The Density Profiles of Massive, Relaxed Galaxy Clusters. II. Separating Luminous and Dark Matter in Cluster Cores
- The Density Profiles of Massive, Relaxed Galaxy Clusters. I. The Total Density Over Three Decades in Radius
- Shear and Ellipticity in Gravitational Lenses
- Fast Automated Analysis of Strong Gravitational Lenses with Convolutional Neural Networks
- Two-dimensional kinematics of SLACS lenses: III. Mass structure and dynamics of early-type lens galaxies beyond z ~ 0.1
- STRIDES: a 3.9 per cent measurement of the Hubble constant from the strong lens system DES J0408-5354
- Fast likelihood-free cosmology with neural density estimators and active learning
- Validating Bayesian Inference Algorithms with Simulation-Based Calibration
- The Sloan Lens ACS Survey. VI: Discovery and analysis of a double Einstein ring
- Galaxy Zoo DECaLS: Detailed Visual Morphology Measurements from Volunteers and Deep Learning for 314,000 Galaxies
- lenstronomy II: A gravitational lensing software ecosystem
- Detections of Lyman Continuum from Star-forming Galaxies at z~3 Through Subaru/Suprime-Cam Narrow-band Imaging
- Flipout: Efficient Pseudo-Independent Weight Perturbations on Mini-Batches
- Rest-Frame Optical Spectra of Three Strongly Lensed Galaxies at z~2
- A Comprehensive guide to Bayesian Convolutional Neural Network with Variational Inference
- Constraining Effective Field Theories with Machine Learning
- Fast calculation of a family of elliptical mass gravitational lens models
- A Guide to Constraining Effective Field Theories with Machine Learning
- The Ultraviolet Spectrum of the Gravitationally Lensed Galaxy `The Cosmic Horseshoe': A Close-up of a Star-forming Galaxy at z = 2
- Cosmological Constraints from the double source plane lens SDSSJ0946+1006
- Uncertainties in Parameters Estimated with Neural Networks: Application to Strong Gravitational Lensing
- Finding strong lenses in CFHTLS using convolutional neural networks
- Keck Spectroscopy of Gravitationally Lensed z=4 Galaxies: Improved Constraints on the Escape Fraction of Ionizing Photons
- Gravitationally Lensed Galaxies at 2<z<3.5: Direct Abundance Measurements of Lya Emitters
- The SL2S Galaxy-scale Lens Sample. II. Cosmic evolution of dark and luminous mass in early-type galaxies
- AutoLens: Automated Modeling of a Strong Lens's Light, Mass and Source
- Automatic physical inference with information maximising neural networks
- Approximate Bayesian Computation for Astronomical Model Analysis: A Case Study in Galaxy Demographics and Morphological Transformation at High Redshift
- The Physical Conditions, Metallicity and Metal Abundance Ratios In a Highly Magnified Galaxy at z = 3.6252
- The Euclid mission design
- Rest-frame ultraviolet spectrum of the gravitationally lensed galaxy `the 8 o'clock arc': stellar and interstellar medium properties
- Testing metallicity indicators at z~1.4 with the gravitationally lensed galaxy CASSOWARY 20
- Mining for Dark Matter Substructure: Inferring subhalo population properties from strong lenses with machine learning
- A measurement of the Hubble constant from angular diameter distances to two gravitational lenses
- Assessing the effect of lens mass model in cosmological application with updated galaxy-scale strong gravitational lensing sample
- Time-delay Cosmography: Increased Leverage with Angular Diameter Distances
- astroABC: An Approximate Bayesian Computation Sequential Monte Carlo sampler for cosmological parameter estimation
- Constraining the dark energy equation of state with double source plane strong lenses
- First direct metallicity measurement of a lensed star-forming galaxy at z=1.7
- Parameters Estimation for the Cosmic Microwave Background with Bayesian Neural Networks
- Turning Back the Clock: Inferring the History of the Eight O'clock Arc
- Line-of-sight effects in strong lensing: Putting theory into practice
- Hierarchical Inference With Bayesian Neural Networks: An Application to Strong Gravitational Lensing
- Project Dinos I: A joint lensing-dynamics constraint on the deviation from the power law in the mass profile of massive ellipticals
- From Images to Dark Matter: End-To-End Inference of Substructure From Hundreds of Strong Gravitational Lenses
- Simulation-Based Calibration Checking for Bayesian Computation: The Choice of Test Quantities Shapes Sensitivity
- A gravitationally-boosted MUSE survey for emission-line galaxies at z>~5 behind the massive cluster RCS 0224
- A Trust Crisis In Simulation-Based Inference? Your Posterior Approximations Can Be Unfaithful
- Ambiguities in gravitational lens models: the density field from the source position transformation
- Large-Scale Gravitational Lens Modeling with Bayesian Neural Networks for Accurate and Precise Inference of the Hubble Constant
- Estimating the warm dark matter mass from strong lensing images with truncated marginal neural ratio estimation
- Observation and Confirmation of Six Strong Lensing Systems in The Dark Energy Survey Science Verification Data
- Strong lens modelling: comparing and combining Bayesian neural networks and parametric profile fitting
- Models of the strongly lensed quasar DES J0408-5354
- Core or cusps: The central dark matter profile of a redshift one strong lensing cluster with a bright central image
- Predictions for Strong Lens Detections with the Nancy Grace Roman Space Telescope
- Truncated proposals for scalable and hassle-free simulation-based inference
- Neural networks and standard cosmography with newly calibrated high redshift GRB observations
- Bayesian model comparison for simulation-based inference
- One never walks alone: the effect of the perturber population on subhalo measurements in strong gravitational lenses
- Identification of Galaxy-Galaxy Strong Lens Candidates in the DECam Local Volume Exploration Survey Using Machine Learning
- Accurate X-ray Timing in the Presence of Systematic Biases With Simulation-Based Inference
- A Systematic Review of Strong Gravitational Lens Modeling Software
- 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
- deeplenstronomy: A dataset simulation package for strong gravitational lensing
- Deep Learning in Wide-field Surveys: Fast Analysis of Strong Lenses in Ground-based Cosmic Experiments
- Hierarchical Neural Simulation-Based Inference Over Event Ensembles
- HNPE: Leveraging Global Parameters for Neural Posterior Estimation
- Modeling Strong Lenses from Wide-Field Ground-Based Observations in KiDS and GAMA
- Simulation-Based Inference of Strong Gravitational Lensing Parameters
- Recommendations for Early Definition Science with the Nancy Grace Roman Space Telescope
- Inferring Structural Parameters of Low-Surface-Brightness-Galaxies with Uncertainty Quantification using Bayesian Neural Networks
- Domain-Adaptive Neural Posterior Estimation for Strong Gravitational Lens Analysis
- Neural Network Prediction of Strong Lensing Systems with Domain Adaptation and Uncertainty Quantification