A Deep Learning Approach for Active Anomaly Detection of Extragalactic Transients

Abstract

There is a shortage of multi-wavelength and spectroscopic followup capabilities given the number of transient and variable astrophysical events discovered through wide-field, optical surveys such as the upcoming Vera C. Rubin Observatory. From the haystack of potential science targets, astronomers must allocate scarce resources to study a selection of needles in real time. Here we present a variational recurrent autoencoder neural network to encode simulated Rubin Observatory extragalactic transient events using 1% of the PLAsTiCC dataset to train the autoencoder. Our unsupervised method uniquely works with unlabeled, real time, multivariate and aperiodic data. We rank 1,129,184 events based on an anomaly score estimated using an isolation forest. We find that our pipeline successfully ranks rarer classes of transients as more anomalous. Using simple cuts in anomaly score and uncertainty, we identify a pure (~95% pure) sample of rare transients (i.e., transients other than Type Ia, Type II and Type Ibc supernovae) including superluminous and pair-instability supernovae. Finally, our algorithm is able to identify these transients as anomalous well before peak, enabling real-time follow up studies in the era of the Rubin Observatory.

arXiv comment13 pages,9 figures, submitted to AAS Journals. Comments welcome!

References in corpus (49)

• Scikit-learn: Machine Learning in Python
• SciPy 1.0--Fundamental Algorithms for Scientific Computing in Python
• Array Programming with NumPy
• Binary companions of evolved stars in APOGEE DR14: Search method and catalog of ~5,000 companions
• LSST: from Science Drivers to Reference Design and Anticipated Data Products
• Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation
• The Zwicky Transient Facility: System Overview, Performance, and First Results
• Improved cosmological constraints from a joint analysis of the SDSS-II and SNLS supernova samples
• Origin of the heavy elements in binary neutron-star mergers from a gravitational wave event
• Deep Learning for Anomaly Detection: A Survey
• SALT2: using distant supernovae to improve the use of Type Ia supernovae as distance indicators
• Nearby Supernova Rates from the Lick Observatory Supernova Search. II. The Observed Luminosity Functions and Fractions of Supernovae in a Complete Sample
• The Pan-STARRS1 Surveys
• An Open Catalog for Supernova Data
• Characterizing the V-band light-curves of hydrogen-rich type II supernovae
• SN 2002cx: The Most Peculiar Known Type Ia Supernova
• K-corrections and Extinction Corrections for Type Ia Supernovae
• The Zwicky Transient Facility Bright Transient Survey. II. A Public Statistical Sample for Exploring Supernova Demographics
• The magnetar model for Type I superluminous supernovae I: Bayesian analysis of the full multicolour light curve sample with MOSFiT
• Pair Instability Supernovae: Light Curves, Spectra, and Shock Breakout
• Weighing Black Holes using Tidal Disruption Events
• MOSFiT: Modular Open-Source Fitter for Transients
• Calcium-rich gap transients in the remote outskirts of galaxies
• The Data Release of the Sloan Digital Sky Survey-II Supernova Survey
• The Automatic Learning for the Rapid Classification of Events (ALeRCE) Alert Broker
• RAPID: Early Classification of Explosive Transients using Deep Learning
• Models and Simulations for the Photometric LSST Astronomical Time Series Classification Challenge (PLAsTiCC)
• LOSS Revisited - II: The relative rates of different types of supernovae vary between low- and high-mass galaxies
• Alert Classification for the ALeRCE Broker System: The Light Curve Classifier
• Understanding extreme quasar optical variability with CRTS: I. Major AGN flares
• Avocado: Photometric Classification of Astronomical Transients with Gaussian Process Augmentation
• UBVRIz Light Curves of 51 Type II Supernovae
• Quasar Selection Based on Photometric Variability
• The Prospects of Observing Tidal Disruption Events with the LSST
• SuperRAENN: A Semi-supervised Supernova Photometric Classification Pipeline Trained on Pan-STARRS1 Medium Deep Survey Supernovae
• Theoretical Models of Optical Transients. I. A Broad Exploration of the Duration-Luminosity Phase Space
• Photometric Redshifts with the LSST: Evaluating Survey Observing Strategies
• The ANTARES Astronomical Time-Domain Event Broker
• Supernova Photometric Classification Pipelines Trained on Spectroscopically Classified Supernovae from the Pan-STARRS1 Medium-Deep Survey
• Anomaly detection in the Zwicky Transient Facility DR3
• Superluminous Supernovae in LSST: Rates, Detection Metrics, and Light Curve Modeling
• PELICAN: deeP architecturE for the LIght Curve ANalysis
• Active Anomaly Detection for time-domain discoveries
• Anomaly Detection in the Open Supernova Catalog
• Anomaly Detection in Gravitational Waves data using Convolutional AutoEncoders
• Photometric Classification of 2315 Pan-STARRS1 Supernovae with Superphot
• Classifying Image Sequences of Astronomical Transients with Deep Neural Networks
• Extended Self-similar Solution for Circumstellar Material-Supernova Ejecta Interaction
• 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

Cited by in corpus (26)

• The DAWES review 10: The impact of deep learning for the analysis of galaxy surveys
• Searching for changing-state AGNs in massive datasets -- I: applying deep learning and anomaly detection techniques to find AGNs with anomalous variability behaviours
• ParSNIP: Generative Models of Transient Light Curves with Physics-Enabled Deep Learning
• Anomaly detection in Hyper Suprime-Cam galaxy images with generative adversarial networks
• Machine Learning for Observational Cosmology
• Learning to detect RFI in radio astronomy without seeing it
• Real-Time Detection of Anomalies in Large-Scale Transient Surveys
• Deep Attention-Based Supernovae Classification of Multi-Band Light-Curves
• Inferencing Progenitor and Explosion Properties of Evolving Core-collapse Supernovae from Zwicky Transient Facility Light Curves
• The ROAD to discovery: machine learning-driven anomaly detection in radio astronomy spectrograms
• Identifying Light-curve Signals with a Deep Learning Based Object Detection Algorithm. II. A General Light Curve Classification Framework
• Searching for Anomalies in the ZTF Catalog of Periodic Variable Stars
• Distance determination of molecular clouds in the 1st quadrant of the Galactic plane using deep learning : I. Method and Results
• Understanding of the properties of neural network approaches for transient light curve approximations
• SNAD Transient Miner: Finding Missed Transient Events in ZTF DR4 using k-D trees
• A Classifier-Based Approach to Multi-Class Anomaly Detection for Astronomical Transients
• Representation Learning for Time-Domain High-Energy Astrophysics: Discovery of Extragalactic Fast X-ray Transient XRT 200515
• ORACLE: A Real-Time, Hierarchical, Deep-Learning Photometric Classifier for the LSST
• A Deep Neural Network Based Reverse Radio Spectrogram Search Algorithm
• Finding radio transients with anomaly detection and active learning based on volunteer classifications
• Scalable and Robust Multiband Modeling of AGN Light Curves in Rubin-LSST
• A machine learning approach to assessing the presence of substructure in quasar host galaxies using the Hyper Suprime-Cam Subaru Strategic Program
• Contaminating Electromagnetic Transients in LISA Gravitational Wave Localization Volumes. I: The Intrinsic Rates
• Quantum Machine Learning for Identifying Transient Events in X-ray Light Curves
• Identifying Anomalous DESI Galaxy Spectra with a Variational Autoencoder
• Predicting large scale cosmological structure evolution with generative adversarial network-based autoencoders