Machine-learning-based Brokers for Real-time Classification of the LSST Alert Stream

Abstract

The unprecedented volume and rate of transient events that will be discovered by the Large Synoptic Survey Telescope (LSST) demands that the astronomical community update its followup paradigm. Alert-brokers -automated software system to sift through, characterize, annotate and prioritize events for followup -will be critical tools for managing alert streams in the LSST era. The Arizona-NOAO Temporal Analysis and Response to Events System (ANTARES) is one such broker. In this work, we develop a machine learning pipeline to characterize and classify variable and transient sources only using the available multiband optical photometry. We describe three illustrative stages of the pipeline, serving the three goals of early, intermediate and retrospective classification of alerts. The first takes the form of variable vs transient categorization, the second, a multi-class typing of the combined variable and transient dataset, and the third, a purity-driven subtyping of a transient class. While several similar algorithms have proven themselves in simulations, we validate their performance on real observations for the first time. We quantitatively evaluate our pipeline on sparse, unevenly sampled, heteroskedastic data from various existing observational campaigns, and demonstrate very competitive classification performance. We describe our progress towards adapting the pipeline developed in this work into a real-time broker working on live alert streams from time-domain surveys.

ROME/REA: A Gravitational Microlensing Search for Exoplanets Beyond the Snow Line on a Global Network of Robotic Telescopes

“…They introduce new targets, adjust relative priorities between targets, alter the observing cadence and exposure times, always using the latest available information. Such intelligent robotic observing agents are of great interest to the wider transient astronomical community as new 19 https://github.com/ebachelet/pyLIMA 20 https://github.com/muLAn-project/ methods are urgently sought to handle the high-volume alert streams from upcoming wide-field surveys like the LSST (Bloom et al 2012;Narayan et al 2018).…”

Surveying the reach and maturity of machine learning and artificial intelligence in astronomy

“…Machine learning is fundamental to accelerating and enabling the data analysis and candidate identification (and rejection) workflows of the ZTF. ML is being used to perform time critical tasks such as morphological star/galaxy classification (Tachibana & Miller, ), binary real/bogus classification of candidates, and asteroid detection (Mahabal et al, ), and can play a role in the brokering of alerts with application to the LSST Alert Stream (Narayan et al, ). In a radio‐based transient object project, Farah et al () used a RF as part of the UTMOST real‐time detection pipeline, leading to the discovery of Fast Radio Burst FRB170827, however, Connor and van Leeuwen () determined that CNNs were sub‐optimal for some radio transient tasks, such as reducing the need for GPU‐accelerated, brute‐force dedispersion of time series signals. Galaxies .…”

Enabling real-time multi-messenger astrophysics discoveries with deep learning

“…The centralized database could be coupled with classification algorithms and catalogs cross-matching. In this sense, the infrastructure of the LSST data management system and the alert brokers under development are good examples 100,101 . The automation of telescopes is an important part of maximizing the number of observed MMA events (for example the Astronomical Event Observatory Network program will provide programmable access to a number of telescopes 102 ).…”