Open Access
Translator Disclaimer
September 2019 Approximate inference for constructing astronomical catalogs from images
Jeffrey Regier, Andrew C. Miller, David Schlegel, Ryan P. Adams, Jon D. McAuliffe, Prabhat
Ann. Appl. Stat. 13(3): 1884-1926 (September 2019). DOI: 10.1214/19-AOAS1258


We present a new, fully generative model for constructing astronomical catalogs from optical telescope image sets. Each pixel intensity is treated as a random variable with parameters that depend on the latent properties of stars and galaxies. These latent properties are themselves modeled as random. We compare two procedures for posterior inference. One procedure is based on Markov chain Monte Carlo (MCMC) while the other is based on variational inference (VI). The MCMC procedure excels at quantifying uncertainty, while the VI procedure is 1000 times faster. On a supercomputer, the VI procedure efficiently uses 665,000 CPU cores to construct an astronomical catalog from 50 terabytes of images in 14.6 minutes, demonstrating the scaling characteristics necessary to construct catalogs for upcoming astronomical surveys.


Download Citation

Jeffrey Regier. Andrew C. Miller. David Schlegel. Ryan P. Adams. Jon D. McAuliffe. Prabhat. "Approximate inference for constructing astronomical catalogs from images." Ann. Appl. Stat. 13 (3) 1884 - 1926, September 2019.


Received: 1 February 2018; Revised: 1 April 2019; Published: September 2019
First available in Project Euclid: 17 October 2019

zbMATH: 07145979
MathSciNet: MR4019161
Digital Object Identifier: 10.1214/19-AOAS1258

Keywords: Astronomy , Graphical model , high performance computing , MCMC , variational inference

Rights: Copyright © 2019 Institute of Mathematical Statistics


Vol.13 • No. 3 • September 2019
Back to Top