Weak-lensing shear measurement with machine learning: teaching artificial neural networks about feature noise

Abstract

Cosmic shear is a primary cosmological probe for several present and upcoming surveys investigating dark matter and dark energy, such as Euclid or WFIRST. The probe requires an extremely accurate measurement of the shapes of millions of galaxies based on imaging data. Crucially, the shear measurement must address and compensate for a range of interwoven nuisance effects related to the instrument optics and detector, noise, unknown galaxy morphologies, colors, blending of sources, and selection effects. This paper explores the use of supervised machine learning (ML) as a tool to solve this inverse problem. We present a simple architecture that learns to regress shear point estimates and weights via shallow artificial neural networks. The networks are trained on simulations of the forward observing process, and take combinations of moments of the galaxy images as inputs. A challenging peculiarity of this ML application is the combination of the noisiness of the input features and the requirements on the accuracy of the inverse regression. To address this issue, the proposed training algorithm minimizes bias over multiple realizations of individual source galaxies, reducing the sensitivity to properties of the overall sample of source galaxies. Importantly, an observational selection function of these source galaxies can be straightforwardly taken into account via the weights. We first introduce key aspects of our approach using toy-model simulations, and then demonstrate its potential on images mimicking Euclid data. Finally, we analyze images from the GREAT3 challenge, obtaining competitively low shear biases despite the use of a simple training set. We conclude that the further development of ML approaches is of high interest to meet the stringent requirements on the shear measurement in current and future surveys. A demonstration implementation of our technique is publicly available.

31 pages, 26 figures, minor changes to match the version published in A&A, code available at https://astro.uni-bonn.de/~mtewes/ml-shear-meas/

References in corpus (12)

• The Shear TEsting Programme 2: Factors affecting high precision weak lensing analyses
• Weak lensing for precision cosmology
• Photometric redshift estimation via deep learning
• Image Analysis for Cosmology: Results from the GREAT10 Galaxy Challenge
• Means of confusion: how pixel noise affects shear estimates for weak gravitational lensing
• Metacalibration: Direct Self-Calibration of Biases in Shear Measurement
• Deep learning for galaxy surface brightness profile fitting
• Photometric redshifts for the Kilo-Degree Survey. Machine-learning analysis with artificial neural networks
• Shear Recovery Accuracy in Weak Lensing Analysis with Elliptical Gauss-Laguerre Method
• A highly precise shear bias estimator independent of the measured shape noise
• Measuring a Charge-Coupled Device Point Spread Function: Euclid Visible Instrument CCD273-84 PSF Performance
• A fast empirical method for galaxy shape measurements in weak lensing surveys

Cited by in corpus (23)

• Euclid. I. Overview of the Euclid mission
• Euclid Preparation IV. Impact of undetected galaxies on weak-lensing shear measurements
• Testing Shear Recovery with Field Distortion
• Baryon acoustic oscillations reconstruction using convolutional neural networks
• Constraining the masses of high-redshift clusters with weak lensing: Revised shape calibration testing for the impact of stronger shears and increased blending
• Cosmic Velocity Field Reconstruction Using AI
• Comparison of Observed Galaxy Properties with Semianalytic Model Predictions using Machine Learning
• Galaxy shape measurement with convolutional neural networks
• Accounting for object detection bias in weak gravitational lensing studies
• Realistic galaxy images and improved robustness in machine learning tasks from generative modelling
• Machine Learning and Cosmology
• Euclid preparation. LIII. LensMC, weak lensing cosmic shear measurement with forward modelling and Markov Chain Monte Carlo sampling
• Noise from Undetected Sources in Dark Energy Survey Images
• A fully data-driven algorithm for accurate shear estimation
• Forklens: Accurate weak-lensing shear measurement with deep learning
• Accurately Estimating Redshifts from CSST Slitless Spectroscopic Survey using Deep Learning
• kSZ Pairwise Velocity Reconstruction with Machine Learning
• Euclid preparation. LXVII. Deep learning true galaxy morphologies for weak lensing shear bias calibration
• GalaxyGenius: Mock galaxy image generator for various telescopes from hydrodynamical simulations
• Clusternets: A deep learning approach to probe clustering dark energy
• Impact of PSF misestimation and galaxy population bias on precision shear measurement using a CNN
• Euclid: Early Release Observations. Weak gravitational lensing analysis of Abell 2390
• D${}_4$CNN$\times$AnaCal: Physics-Informed Machine Learning for Accurate and Precise Weak Lensing Shear Estimation