Searching for changing-state AGNs in massive datasets -- I: applying deep learning and anomaly detection techniques to find AGNs with anomalous variability behaviours

Abstract

The classic classification scheme for Active Galactic Nuclei (AGNs) was recently challenged by the discovery of the so-called changing-state (changing-look) AGNs (CSAGNs). The physical mechanism behind this phenomenon is still a matter of open debate and the samples are too small and of serendipitous nature to provide robust answers. In order to tackle this problem, we need to design methods that are able to detect AGN right in the act of changing-state. Here we present an anomaly detection (AD) technique designed to identify AGN light curves with anomalous behaviors in massive datasets. The main aim of this technique is to identify CSAGN at different stages of the transition, but it can also be used for more general purposes, such as cleaning massive datasets for AGN variability analyses. We used light curves from the Zwicky Transient Facility data release 5 (ZTF DR5), containing a sample of 230,451 AGNs of different classes. The ZTF DR5 light curves were modeled with a Variational Recurrent Autoencoder (VRAE) architecture, that allowed us to obtain a set of attributes from the VRAE latent space that describes the general behaviour of our sample. These attributes were then used as features for an Isolation Forest (IF) algorithm, that is an anomaly detector for a "one class" kind of problem. We used the VRAE reconstruction errors and the IF anomaly score to select a sample of 8,809 anomalies. These anomalies are dominated by bogus candidates, but we were able to identify 75 promising CSAGN candidates.

arXiv commentAccepted for publication in the Astronomical Journal (AJ)

References in corpus (56)

• The NumPy array: a structure for efficient numerical computation
• The Sloan Digital Sky Survey: Technical Summary
• LSST: from Science Drivers to Reference Design and Anticipated Data Products
• X-ray Properties of Black-Hole Binaries
• The SIMBAD astronomical database
• The Zwicky Transient Facility: System Overview, Performance, and First Results
• Towards a unified model for black hole X-ray binary jets
• Modelling the behaviour of accretion flows in X-ray binaries
• First Results from the Catalina Real-time Transient Survey
• The Zwicky Transient Facility: Data Processing, Products, and Archive
• Are the Variations in Quasar Optical Flux Driven by Thermal Fluctuations?
• Modeling the Time Variability of SDSS Stripe 82 Quasars as a Damped Random Walk
• Active Galactic Nuclei as scaled-up Galactic black holes
• Active Galactic Nuclei: what's in a name?
• Studies in Astronomical Time Series Analysis. VI. Bayesian Block Representations
• The Ensemble Photometric Variability of ~25000 Quasars in the Sloan Digital Sky Survey
• The Discovery of the First "Changing Look" Quasar: New Insights into the Physics & Phenomenology of AGN
• The 5th edition of the Roma-BZCAT. A short presentation
• A Systematic Search for Changing-Look Quasars in SDSS
• Spectral properties of quasars from Sloan Digital Sky Survey data release 14: The catalog
• The Automatic Learning for the Rapid Classification of Events (ALeRCE) Alert Broker
• 1ES 1927+654: An AGN Caught Changing Look on a Timescale of Months
• The Half Million Quasars (HMQ) Catalogue
• A Mid-IR Selected Changing-Look Quasar and Physical Scenarios for Abrupt AGN Fading
• Changing-Look Quasar Candidates: First Results from Follow-up Spectroscopy of Highly Optically Variable Quasars
• The destruction and recreation of the X-ray corona in a changing-look Active Galactic Nucleus
• A Catalog of Narrow Line Seyfert 1 Galaxies from the Sloan Digital Sky Survey Data Release 12
• Alert Classification for the ALeRCE Broker System: The Light Curve Classifier
• A recurrent neural network for classification of unevenly sampled variable stars
• A new physical interpretation of optical and infrared variability in quasars
• Understanding extreme quasar optical variability with CRTS: I. Major AGN flares
• A New Catalogue of Type 1 AGN and its Implication on the AGN Unified Model
• Optical variability of AGN in the PTF/iPTF survey
• The Analogous Structure of Accretion Flows in Supermassive and Stellar Mass Black Holes: New Insights from Faded Changing-Look Quasars
• Understanding extreme quasar optical variability with CRTS: II. Changing-state quasars
• A New Class of Changing-Look LINERs
• Pan-STARRS1 variability of XMM-COSMOS AGN. II. Physical correlations and power spectrum analysis
• A new class of flares from accreting supermassive black holes
• A Morphological Classification Model to Identify Unresolved PanSTARRS1 Sources: Application in the ZTF Real-Time Pipeline
• Improving Damped Random Walk parameters for SDSS Stripe 82 Quasars with Pan-STARRS1
• SuperRAENN: A Semi-supervised Supernova Photometric Classification Pipeline Trained on Pan-STARRS1 Medium Deep Survey Supernovae
• Dimensionality Reduction of SDSS Spectra with Variational Autoencoders
• Astronomaly: Personalised Active Anomaly Detection in Astronomical Data
• Evidence for a change in the X-ray radiation mechanism in the hard state of Galactic black holes
• The QUEST-La Silla AGN variability survey: connection between AGN variability and black hole physical properties
• Anomaly detection in the Zwicky Transient Facility DR3
• A Deep Learning Approach for Active Anomaly Detection of Extragalactic Transients
• Near Infrared Variability of obscured and unobscured X-ray selected AGN in the COSMOS field
• Active Anomaly Detection for time-domain discoveries
• Deep modeling of quasar variability
• Multi-Wavelength Monitoring and Reverberation Mapping of a Changing Look Event in the Seyfert Galaxy NGC 3516
• The ZTF Source Classification Project: I. Methods and Infrastructure
• High-redshift Extreme Variability Quasars from Sloan Digital Sky Survey Multi-Epoch Spectroscopy
• Evidence of Rapid Optical Variability In Selected Narrow-Line Seyfert 1 Galaxies
• Unveiling the physics of low luminosity AGN through X-ray variability: LINER versus Seyfert 2
• Characterization of Optical Light Curves of Extreme Variability Quasars Over a ~16-yr Baseline

Cited by in corpus (24)

• The DAWES review 10: The impact of deep learning for the analysis of galaxy surveys
• Confirming new changing--look AGNs discovered through optical variability using a random-forest based light curve classifier
• The rise and fall of the iron-strong nuclear transient PS16dtm
• The SNAD Viewer: Everything You Want to Know about Your Favorite ZTF Object
• Deep Attention-Based Supernovae Classification of Multi-Band Light-Curves
• Persistent and occasional: searching for the variable population of the ZTF/4MOST sky using ZTF data release 11
• ATAT: Astronomical Transformer for time series And Tabular data
• Long-term optical spectral monitoring of a changing-look active galactic nucleus NGC 3516 -- II. Broad-line profile variability
• Latent Stochastic Differential Equations for Modeling Quasar Variability and Inferring Black Hole Properties
• Supernova search with active learning in ZTF DR3
• Searching for Anomalies in the ZTF Catalog of Periodic Variable Stars
• Conditional Neural Process for non-parametric modeling of AGN light curve
• Unveiling the quasar main sequence: illuminating the complexity of active galactic nuclei and their evolution
• Joint Modeling of Quasar Variability and Accretion Disk Reprocessing using Latent Stochastic Differential Equations
• Real-Time Active Learning for optimised spectroscopic follow-up: Enhancing early SN Ia classification with the Fink broker
• Real-bogus scores for active anomaly detection
• Learning novel representations of variable sources from multi-modal $\text{Gaia}$ data via autoencoders
• Convolutional Deep Denoising Autoencoders for Radio Astronomical Images
• Interpreting anomaly detection of SDSS spectra
• Modeling quasar variability through self-organizing map-based process
• Quantum Machine Learning for Identifying Transient Events in X-ray Light Curves
• Automated all-sky detection of γ Doradus / δ Scuti hybrids in TESS data from positive unlabelled (PU) learning
• Semi-Supervised Learning for Lensed Quasar Detection
• Towards model-free stellar chemical abundances. Potential applications in the search for chemically peculiar stars in large spectroscopic surveys