Skip to main content

Approximate inference for constructing astronomical catalogs from images

Author(s): Regier, Jeffrey; Miller, Andrew C; Schlegel, David; Adams, Ryan P; McAuliffe, Jon D; et al

Download
To refer to this page use: http://arks.princeton.edu/ark:/88435/pr1tz63
Abstract: © Institute of Mathematical Statistics, 2019. 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.
Publication Date: 1-Sep-2019
Citation: Regier, J, Miller, AC, Schlegel, D, Adams, RP, McAuliffe, JD, Prabhat. (2019). Approximate inference for constructing astronomical catalogs from images. Annals of Applied Statistics, 13 (3), 1884 - 1926. doi:10.1214/19-AOAS1258
DOI: doi:10.1214/19-AOAS1258
ISSN: 1932-6157
EISSN: 1941-7330
Pages: 1884 - 1926
Type of Material: Journal Article
Journal/Proceeding Title: Annals of Applied Statistics
Version: Author's manuscript



Items in OAR@Princeton are protected by copyright, with all rights reserved, unless otherwise indicated.