RAPID: Early Classification of Explosive Transients Using Deep Learning

Abstract

We present RAPID (Real-time Automated Photometric IDentification), a novel time-series classification tool capable of automatically identifying transients from within a day of the initial alert, to the full lifetime of a light curve. Using a deep recurrent neural network with Gated Recurrent Units (GRUs), we present the first method specifically designed to provide early classifications of astronomical time-series data, typing 12 different transient classes. Our classifier can process light curves with any phase coverage, and it does not rely on deriving computationally expensive features from the data, making RAPID well-suited for processing the millions of alerts that ongoing and upcoming wide-field surveys such as the Zwicky Transient Facility (ZTF), and the Large Synoptic Survey Telescope (LSST) will produce. The classification accuracy improves over the lifetime of the transient as more photometric data becomes available, and across the 12 transient classes, we obtain an average area under the receiver operating characteristic curve of 0.95 and 0.98 at early and late epochs, respectively. We demonstrate RAPID's ability to effectively provide early classifications of observed transients from the ZTF data stream. We have made RAPID available as an open-source software package a) for machine learning-based alert-brokers to use for the autonomous and quick classification of several thousand light curves within a few seconds.

GROWTH on S190425z: Searching Thousands of Square Degrees to Identify an Optical or Infrared Counterpart to a Binary Neutron Star Merger with the Zwicky Transient Facility and Palomar Gattini-IR

“…Going forward, prioritizing further automatized classification of objects can lead to more rapid follow-up and dissemination of the most interesting objects. For example, the inclusion of machine-learning-based photometric classification codes such as RAPID (Muthukrishna et al 2019) will help facilitate candidate selection and prioritization. We are also actively improving the scheduling optimization, and have since added a Figure 5.…”

The Zwicky Transient Facility Bright Transient Survey. II. A Public Statistical Sample for Exploring Supernova Demographics*

“…Large numbers of transients are of limited scientific value without secure classifications and redshifts (Kulkarni 2020). Despite recent advances in photometric classification (e.g., Muthukrishna et al 2019;Villar et al 2019Villar et al , 2020Dauphin et al 2020;Hosseinzadeh et al 2020), the only ground truth for this remains spectroscopy, an observationally expensive endeavor. Deciding which transients to spectroscopically classify and which to ignore typically involves extensive human decision-making, potentially introducing complex biases and diminishing the value of large statistical samples for studies of (for example) volumetric rates, luminosity functions, or ensemble host-galaxy properties.…”

Supernova host galaxies in the dark energy survey: I. Deep coadds, photometry, and stellar masses

“…As transient surveys probe larger areas with deeper observations, however, it not feasible to classify all of the SNe spectroscopically. We thus define samples by classifying SNe 'photometrically', principally using the light-curve shape and colour to distinguish SNe Ia from core-collapse events using classifiers such as pSNid (Sako et al 2008), SuperNNova (Möller & de Boissière 2019), and RAPID (Muthukrishna et al 2019).…”