Automating Discovery and Classification of Transients and Variable Stars in the Synoptic Survey Era

Abstract

The rate of image acquisition in modern synoptic imaging surveys has already begun to outpace the feasibility of keeping astronomers in the real-time discovery and classification loop. Here we present the inner workings of a framework, based on machine-learning algorithms, that captures expert training and ground-truth knowledge about the variable and transient sky to automate 1) the process of discovery on image differences and, 2) the generation of preliminary science-type classifications of discovered sources. Since follow-up resources for extracting novel science from fast-changing transients are precious, self-calibrating classification probabilities must be couched in terms of efficiencies for discovery and purity of the samples generated. We estimate the purity and efficiency in identifying real sources with a two-epoch image-difference discovery algorithm for the Palomar Transient Factory (PTF) survey. Once given a source discovery, using machine-learned classification trained on PTF data, we distinguish between transients and variable stars with a 3.8% overall error rate (with 1.7% errors for imaging within the Sloan Digital Sky Survey footprint). At >96% classification efficiency, the samples achieve 90% purity. Initial classifications are shown to rely primarily on context-based features, determined from the data itself and external archival databases. In the ~one year since autonomous operations, this discovery and classification framework has led to several significant science results, from outbursting young stars to subluminous Type IIP supernovae to candidate tidal disruption events. We discuss future directions of this approach, including the possible roles of crowdsourcing and the scalability of machine learning to future surveys such a the Large Synoptical Survey Telescope (LSST).

43 pages, 13 figures, submitted to Publications of the Astronomical Society of the Pacific (PASP)

References in corpus (20)

• Improved background subtraction for the Sloan Digital Sky Survey images
• Observed Fractions of Core-Collapse Supernova Types and Initial Masses of their Single and Binary Progenitor Stars
• Supernova 2011fe from an Exploding Carbon-Oxygen White Dwarf Star
• Constraints on the Progenitor System of the Type Ia Supernova SN 2011fe/PTF11kly
• On Machine-Learned Classification of Variable Stars with Sparse and Noisy Time-Series Data
• A Compact Degenerate Primary-Star Progenitor of SN 2011fe
• The Automated Palomar 60-Inch Telescope
• Automated supervised classification of variable stars I. Methodology
• Optimal Time-Series Selection of Quasars
• Real-Time Detection and Rapid Multiwavelength Follow-up Observations of a Highly Subluminous Type II-P Supernova from the Palomar Transient Factory Survey
• Evidence for an FU Orionis-like Outburst from a Classical T Tauri Star
• The Subluminous and Peculiar Type Ia Supernova PTF09dav
• How to Find More Supernovae with Less Work: Object Classification Techniques for Difference Imaging
• PTF10nvg: An Outbursting Class I Protostar in the Pelican/North American Nebula
• Active Learning to Overcome Sample Selection Bias: Application to Photometric Variable Star Classification
• Automated supervised classification of variable stars II. Application to the OGLE database
• Hubble Space Telescope Studies of Nearby Type Ia Supernovae: The Mean Maximum Light Ultraviolet Spectrum and its Dispersion
• Towards Real-time Classification of Astronomical Transients
• An Autonomous Adaptive Scheduling Agent for Period Searching
• SDSS, LSST, and Gaia: Lessons and Synergies

Cited by in corpus (120)

• Illuminating Gravitational Waves: A Concordant Picture of Photons from a Neutron Star Merger
• A Continuum of H- to He-Rich Tidal Disruption Candidates With a Preference for E+A Galaxies
• Proper image subtraction - optimal transient detection, photometry and hypothesis testing
• The Automatic Learning for the Rapid Classification of Events (ALeRCE) Alert Broker
• Probabilistic Random Forest: A machine learning algorithm for noisy datasets
• PTF11iqb: Cool supergiant mass loss that bridges the gap between Type IIn and normal supernovae
• Automated Transient Identification in the Dark Energy Survey
• RAPID: Early Classification of Explosive Transients using Deep Learning
• Science with a wide-field UV transient explorer
• The Host Galaxies of Type Ia Supernovae Discovered by the Palomar Transient Factory
• The weirdest SDSS galaxies: results from an outlier detection algorithm
• The Caltech-NRAO Stripe 82 Survey (CNSS) Paper I: The Pilot Radio Transient Survey In 50 deg${}^2$
• Using Machine Learning for Discovery in Synoptic Survey Imaging
• Deep-HiTS: Rotation Invariant Convolutional Neural Network for Transient Detection
• Searching for Electromagnetic Counterparts to Gravitational-wave Merger Events with the Prototype Gravitational-wave Optical Transient Observer (GOTO-4)
• The Young Supernova Experiment: Survey Goals, Overview, and Operations
• Exploring the spectral diversity of low-redshift Type Ia supernovae using the Palomar Transient Factory
• A search for optical AGN variability in 35,000 low-mass galaxies with the Palomar Transient Factory
• A multi-wavelength investigation of the radio-loud supernova PTF11qcj and its circumstellar environment
• Machine Learning-based Brokers for Real-time Classification of the LSST Alert Stream
• The IPAC Image Subtraction and Discovery Pipeline for the intermediate Palomar Transient Factory
• Asteroid lightcurves from the Palomar Transient Factory survey: Rotation periods and phase functions from sparse photometry
• Discovery of a Cosmological, Relativistic Outburst via its Rapidly Fading Optical Emission
• Rapidly Rising Transients from Subaru Hyper Suprime-Cam Transient Survey
• Observational constraints on the optical and near-infrared emission from the neutron star-black hole binary merger S190814bv
• Autoregressive Times Series Methods for Time Domain Astronomy
• The LSST DESC DC2 Simulated Sky Survey
• Machine learning for transient discovery in Pan-STARRS1 difference imaging
• Realfast: Real-Time, Commensal Fast Transient Surveys with the Very Large Array
• First Searches for Optical Counterparts to Gravitational-wave Candidate Events
• Highly Variable Extinction and Accretion in the Jet-driving Class I Type Young Star PTF 10nvg (V2492 Cyg, IRAS 20496+4354)
• Discovery and redshift of an optical afterglow in 71 square degrees: iPTF13bxl and GRB 130702A
• The volumetric rate of normal type Ia supernovae in the local universe discovered by the Palomar Transient Factory
• Type Ia Supernova Spectral Features in the Context of Their Host Galaxy Properties
• Machine Learning Classification of SDSS Transient Survey Images
• A transient search using combined human and machine classifications
• Combining machine learning techniques, microanalyses and large geochemical datasets for tephrochronological studies in complex volcanic areas: new age constraints for the Pleistocene magmatism of Central Italy
• Automated Classification of Periodic Variable Stars detected by the Wide-field Infrared Survey Explorer
• Deep Learning for Image Sequence Classification of Astronomical Events
• Characterizing the local relation between star formation rate and gas-phase metallicity in MaNGA spiral galaxies
• Machine learning technique for morphological classification of galaxies from the SDSS. I. Photometry-based approach
• Convolutional Neural Networks for Transient Candidate Vetting in Large-Scale Surveys
• Detecting outliers and learning complex structures with large spectroscopic surveys - a case study with APOGEE stars
• Kiso Supernova Survey (KISS): Survey Strategy
• Transient-optimised real-bogus classification with Bayesian Convolutional Neural Networks -- sifting the GOTO candidate stream
• Small near-Earth asteroids in the Palomar Transient Factory survey: A real-time streak-detection system
• A machine-learning method for identifying multi-wavelength counterparts of submillimeter galaxies: training and testing using AS2UDS and ALESS
• Unsupervised machine learning for transient discovery in Deeper, Wider, Faster light curves
• Machine Learning for Observational Cosmology
• Clustering Based Feature Learning on Variable Stars
• The ZTF Source Classification Project: I. Methods and Infrastructure
• Unsupervised Classification of Variable Stars
• Machine-learning Selection of Optical Transients in Subaru/Hyper Suprime-Cam Survey
• Euclid: Superluminous supernovae in the Deep Survey
• The SkyMapper Transient Survey
• Light curve classification with recurrent neural networks for GOTO: dealing with imbalanced data
• Automatic classification of time-variable X-ray sources
• Color Me Intrigued: the Discovery of iPTF 16fnm, a Supernova 2002cx-like Object
• Discovery of New Dipper Stars with K2: A Window into the Inner Disk Region of T Tauri Stars
• J-GEM Follow-Up Observations to Search for an Optical Counterpart of The First Gravitational Wave Source GW150914
• J-GEM Follow-Up Observations of The Gravitational Wave Source GW151226
• Real-Time Recovery Efficiencies and Performance of the Palomar Transient Factory's Transient Discovery Pipeline
• iPTF Archival Search for Fast Optical Transients
• Multivariate Classification with Random Forests for Gravitational Wave Searches of Black Hole Binary Coalescence
• Deep Neural Network Classifier for Variable Stars with Novelty Detection Capability
• The Variable Sky of Deep Synoptic Surveys
• A Machine Learning Classifier for Microlensing in Wide-Field Surveys
• Vetting the optical transient candidates detected by the GWAC network using convolutional neural networks
• The Photometric LSST Astronomical Time-series Classification Challenge (PLAsTiCC): Selection of a performance metric for classification probabilities balancing diverse science goals
• Classifying Image Sequences of Astronomical Transients with Deep Neural Networks
• Real-Time Data Mining of Massive Data Streams from Synoptic Sky Surveys
• Machine Learning on Difference Image Analysis: A comparison of methods for transient detection
• The High Cadence Transient Survey (HITS): Compilation and characterization of light-curve catalogs
• Machine Learning Based Real Bogus System for HSC-SSP Moving Object Detecting Pipeline
• Beyond the Local Volume. I. Surface Densities of Ultracool Dwarfs in Deep HST/WFC3 Parallel Fields
• Preparing for advanced LIGO: A Star-Galaxy Separation Catalog for the Palomar Transient Factory
• Statistical Searches for Microlensing Events in Large, Non-Uniformly Sampled Time-Domain Surveys: A Test Using Palomar Transient Factory Data
• The VVV Infrared Variability Catalog (VIVA-I)
• Modeling Light Curves for Improved Classification
• Machine Learning for Transient Recognition in Difference Imaging With Minimum Sampling Effort
• Uncertain classification of Variable Stars: handling observational GAPS and noise
• ANTARES: A Prototype Transient Broker System
• The HST See Change Program: I. Survey Design, Pipeline, and Supernova Discoveries
• A classification algorithm for time-domain novelties in preparation for LSST alerts: Application to variable stars and transients detected with DECam in the Galactic Bulge
• Single-epoch supernova classification with deep convolutional neural networks
• Density Based Outlier Scoring on Kepler Data
• deepSIP: Linking Type Ia Supernova Spectra to Photometric Quantities with Deep Learning
• The Palomar Transient Factory Sky2Night programme
• X-ray Transients in the Advanced LIGO/Virgo Horizon
• GS-TEC: the Gaia Spectrophotometry Transient Events Classifier
• What's the Difference? The potential for Convolutional Neural Networks for transient detection without template subtraction
• Variable classification in the LSST era: Exploring a model for quasi-periodic light curves
• The radial distribution of supernovae compared to star formation tracers
• Radio Afterglows from Compact Binary Coalescences: Prospects for Next-Generation Telescopes
• MANTRA: A Machine Learning reference lightcurve dataset for astronomical transient event recognition
• iPTF Survey for Cool Transients
• ROME/REA: A gravitational microlensing search for exo-planets beyond the snow-line on a global network of robotic telescopes
• On the optical counterparts of radio transients and variables
• Selection of Burst-like Transients and Stochastic Variables Using Multi-Band Image Differencing in the Pan-STARRS1 Medium-Deep Survey
• The four cosmic tidal web elements from the $\beta$-skeleton
• A novel method for transient detection in high-cadence optical surveys: Its application for a systematic search for novae in M31
• Identifying Exoplanets with Deep Learning. IV. Removing Stellar Activity Signals from Radial Velocity Measurements Using Neural Networks
• Deep-learning Real/Bogus classification for the Tomo-e Gozen transient survey
• Identifying Transients in the Dark Energy Survey using Convolutional Neural Networks
• Self-Supervised Clustering on Image-Subtracted Data with Deep-Embedded Self-Organizing Map
• Automated Transient Detection with Shapelet Analysis in Image-subtracted Data
• Radio view of a broad-line Type Ic supernova ASASSN-16fp
• Life Beyond PTF
• A Morphological Classification Model to Identify Unresolved PanSTARRS1 Sources II: Update to the PS1 Point Source Catalog
• An Algorithm for the Visualization of Relevant Patterns in Astronomical Light Curves
• Multi-scale stamps for real-time classification of alert streams
• Informative Bayesian model selection for RR Lyrae star classifiers
• The Gravitational-wave Optical Transient Observer (GOTO): prototype performance and prospects for transient science
• Large-Scale Image Processing with the ROTSE Pipeline for Follow-Up of Gravitational Wave Events
• The automation of optical transient discovery and classification in Rubin-era time-domain astronomy
• Unsupervised learning for variability detection with Gaia DR3 photometry. The main sequence-white dwarf valley
• NGTS clusters survey IV. Search for Dipper stars in the Orion Nebular Cluster
• Informative regularization for a multi-layer perceptron RR Lyrae classifier under data shift
• Searching for outliers in the Chandra Source Catalog
• Finding Fast Transients in Real Time Using Novel Light Curve Analysis Algorithm