Erratum: “Automated Transient Identification in the Dark Energy Survey” (2015, Aj, 150, 82)

Abstract

We describe an algorithm for identifying point-source transients and moving objects on reference-subtracted optical images containing artifacts of processing and instrumentation. The algorithm makes use of the supervised machine learning technique known as Random Forest. We present results from its use in the Dark Energy Survey Supernova program (DES-SN), where it was trained using a sample of 898,963 signal and background events generated by the transient detection pipeline. After reprocessing the data collected during the first DES-SN observing season (Sep. 2013 through Feb. 2014) using the algorithm, the number of transient candidates eligible for human scanning decreased by a factor of 13.4, while only 1.0 percent of the artificial Type Ia supernovae (SNe) injected into search images to monitor survey efficiency were lost, most of which were very faint events. Here we characterize the algorithm's performance in detail, and we discuss how it can inform pipeline design decisions for future time-domain imaging surveys, such as the Large Synoptic Survey Telescope and the Zwicky Transient Facility.

Rates and Properties of Supernovae Strongly Gravitationally Lensed by Elliptical Galaxies in Time-domain Imaging Surveys

“…The improved spatial resolution (0.2 pixels) of LSST compared to ZTF enables some gLSNe to be totally or marginally resolved, but the majority of systems remain unresolved. LSST must take special care to ensure that its machine learning algorithm for difference image artifact rejection (e.g., Goldstein et al 2015) does not reject marginally resolved gLSNe, such as the ones in row 5, column 1; row 1, column 4; and row 1, column 5.…”

A Deep Learning Approach to Galaxy Cluster X-Ray Masses

“…Machine learning (ML) offers a number of tools that can be used to untangle subtle signals and extract complicated correlations. ML has been utilized in astronomy and cosmology for classification tasks such as labeling galaxy morphology (Banerji et al 2010;Dieleman et al 2015;Domínguez Sánchez et al 2018), identifying transient types (Goldstein et al 2015), identifying the presence or absence of lensing signals in images (Lanusse et al 2018), categorizing the type of sources driving reionization (Hassan et al 2018), and estimating photometric redshifts (Pasquet et al 2018). ML has also been used for astronomical and cosmological regression tasks, for example, reducing errors in cluster dynamical mass measurements (Ntampaka et al 2015(Ntampaka et al , 2016Armitage et al 2018), determining the duration of reionization (La Plante & Ntampaka 2018), and producing tighter cosmological parameter constraints with mock catalogs (Gupta et al 2018).…”

GROWTH on S190426c: Real-time Search for a Counterpart to the Probable Neutron Star–Black Hole Merger using an Automated Difference Imaging Pipeline for DECam

“…We rejected any objects that overlapped masked pixels on either the template or science images, had SExtractor extraction flags, had an axis ratio greater than 1.5, had a FWHM more than twice the seeing, had a PSF magnitude greater than 30, had a signal-to-noise ratio less than 5, or had a semi-major axis less than 1 pixel. After making these initial cuts, we used the publicly available autoScan code (Goldstein et al 2015), based on the machine learning technique Random Forest, to probabilistically classify the "realness" of the remaining extracted sources. The code has been successfully used in past DECam searches for GW counterparts in independent difference imaging pipelines (e.g., Soares-Santos et al 2017b).…”