ABC-SN: Attention Based Classifier for Supernova Spectra

Abstract

While significant advances have been made in photometric classification ahead of the millions of transient events and hundreds of supernovae (SNe) each night that the Vera C. Rubin Observatory Legacy Survey of Space and Time (LSST) will discover, classifying SNe spectroscopically remains the best way to determine most subtypes of SNe. Traditional spectrum classification tools use template matching techniques (eg., SNID, SuperFit) and require significant human supervision. Two deep learning spectral classifiers, DASH and SNIascore, define the state of the art, but SNIascore is a binary classifier devoted to maximizing the purity of the SN Ia-norm sample, while DASH is no longer maintained and the original work suffers from contamination of multi-epoch spectra in the training and test sets. We have explored several neural network architectures in order to create a new automated method for classifying SN subtypes, settling on an attention-based model we call ABC-SN. We benchmark our results against an updated version of DASH, thus providing the community with an up-to-date general-purpose SN classifier. Our dataset is comprised of ten different SN subtypes, including subtypes of SN Ia, core collapse, and interacting SNe. We find that ABC-SN outperforms DASH for nearly all classes, including an improvement of $26\%$ in SN Ia completeness ($\sim 88\%$) and $2.4\%$ in SN Ia purity ($\sim 95\%$) when unthresholded (improvements for each class can further be obtained by tuned thresholds), and we discuss the limitations of current SN datasets for benchmarking performance.

29 pages, 14 figures, 5 tables. Published in The Astrophysical Journal, Volume 1000, Number 1

References in corpus (44)

• WISeREP - An Interactive Supernova Data Repository
• Pulsational Pair-Instability Supernovae
• The type Ia supernova SNLS-03D3bb from a super-Chandrasekhar-mass white dwarf star
• Determining the Type, Redshift, and Age of a Supernova Spectrum
• A Common Explosion Mechanism for Type Ia Supernovae
• Berkeley Supernova Ia Program I: Observations, Data Reduction, and Spectroscopic Sample of 582 Low-Redshift Type Ia Supernovae
• Sub-Chandrasekhar Mass Models For Type Ia Supernovae
• The Physics of Core-Collapse Supernovae
• The Spectroscopic Diversity of Type Ia Supernovae
• The Luminous and Carbon-Rich Supernova 2006gz: A Double Degenerate Merger?
• Pair Instability Supernovae: Light Curves, Spectra, and Shock Breakout
• The SED Machine: a robotic spectrograph for fast transient classification
• Gemini Spectroscopy of Supernovae from SNLS: Improving High Redshift SN Selection and Classification
• Optical Spectra of 73 Stripped-Envelope Core-Collapse Supernovae
• The Spectral SN-GRB Connection: Systematic Spectral Comparisons between Type Ic Supernovae, and broad-lined Type Ic Supernovae with and without Gamma-Ray Bursts
• SuperNNova: an open-source framework for Bayesian, Neural Network based supernova classification
• Observational and Physical Classification of Supernovae
• Type II supernova spectral diversity I: Observations, sample characterization and spectral line evolution
• Superluminous supernovae
• Analyzing the Largest Spectroscopic Dataset of Stripped Supernovae to Improve Their Identifications and Constrain Their Progenitors
• Optical Spectroscopy of Type Ia Supernovae
• Investigating the properties of stripped-envelope supernovae, what are the implications for their progenitors?
• ESC Supernova spectroscopy of non-ESC targets
• The early detection and follow-up of the highly obscured Type II supernova 2016ija/DLT16am
• Avocado: Photometric Classification of Astronomical Transients with Gaussian Process Augmentation
• DASH: Deep Learning for the Automated Spectral Classification of Supernovae and their Hosts
• Luminous Supernovae: Unveiling a Population Between Superluminous and Normal Core-collapse Supernovae
• Optimal Classification and Outlier Detection for Stripped-Envelope Core-Collapse Supernovae
• Exploring the spectroscopic diversity of type Ia supernovae with DRACULA: a machine learning approach
• SN 2018gjx reveals that some SNe Ibn are SNe IIb exploding in dense circumstellar material
• Line Identifications of Type I Supernovae: On the Detection of Si II for these Hydrogen-poor Events
• Modeling of the nebular-phase spectral evolution of stripped-envelope supernovae. New grids from 100 to 450 days
• Broad-emission-line dominated hydrogen-rich luminous supernovae
• Optical studies of two stripped envelope supernovae SN 2015ap (Type Ib) and SN 2016P (Type Ic)
• Astroconformer: The Prospects of Analyzing Stellar Light Curves with Transformer-Based Deep Learning Models
• Origin of high-velocity ejecta and early red excess emission in the infant Type Ia supernova 2021aefx
• SN 2019ewu: A Peculiar Supernova with Early Strong Carbon and Weak Oxygen Features from a New Sample of Young SN Ic Spectra
• Velocity evolution of broad-lined type-Ic supernovae with and without gamma-ray bursts
• Spectral dataset of young type Ib supernovae and their time evolution
• The carbon-rich type Ic supernova 2016adj in the iconic dust lane of Centaurus A: signatures of interaction with circumstellar hydrogen?
• ORACLE: A Real-Time, Hierarchical, Deep-Learning Photometric Classifier for the LSST
• Testing and Combining Transient Spectral Classification Tools on 4MOST-like Blended Spectra
• Super-SNID : an expanded set of SNID classes and templates for the new era of wide-field surveys
• Normal or transitional? The evolution and properties of two type Ia supernovae in the Virgo cluster