The weirdest SDSS galaxies: results from an outlier detection algorithm

Abstract

How can we discover objects we did not know existed within the large datasets that now abound in astronomy? We present an outlier detection algorithm that we developed, based on an unsupervised Random Forest. We test the algorithm on more than two million galaxy spectra from the Sloan Digital Sky Survey and examine the 400 galaxies with the highest outlier score. We find objects which have extreme emission line ratios and abnormally strong absorption lines, objects with unusual continua, including extremely reddened galaxies. We find galaxy-galaxy gravitational lenses, double-peaked emission line galaxies, and close galaxy pairs. We find galaxies with high ionisation lines, galaxies which host supernovae, and galaxies with unusual gas kinematics. Only a fraction of the outliers we find were reported by previous studies that used specific and tailored algorithms to find a single class of unusual objects. Our algorithm is general and detects all of these classes, and many more, regardless of what makes them peculiar. It can be executed on imaging, time-series, and other spectroscopic data, operates well with thousands of features, is not sensitive to missing values, and is easily parallelisable.

MNRAS accepted. Weirdness score for all SDSS galaxies at http://www.wise-obs.tau.ac.il/~dovip/weird-galaxies/ and code available at https://github.com/dalya/WeirdestGalaxies

References in corpus (105)

• Scikit-learn: Machine Learning in Python
• Stellar population synthesis at the resolution of 2003
• The Sloan Digital Sky Survey: Technical Summary
• The Host Galaxies of AGN
• The Origin of the Mass--Metallicity Relation: Insights from 53,000 Star-Forming Galaxies in the SDSS
• The physical properties of star forming galaxies in the low redshift universe
• Correcting for the Effects of Interstellar Extinction
• Theoretical Modeling of Starburst Galaxies
• The Eleventh and Twelfth Data Releases of the Sloan Digital Sky Survey: Final Data from SDSS-III
• Stellar Masses and Star Formation Histories for 10^5 Galaxies from the Sloan Digital Sky Survey
• Modeling feedback from stars and black holes in galaxy mergers
• Parametric Recovery of Line-of-Sight Velocity Distributions from Absorption-Line Spectra of Galaxies via Penalized Likelihood
• The Host Galaxies and Classification of Active Galactic Nuclei
• A Unified, Merger-Driven Model for the Origin of Starbursts, Quasars, the Cosmic X-Ray Background, Supermassive Black Holes and Galaxy Spheroids
• MILES: A Medium resolution INT Library of Empirical Spectra
• Galactic Winds
• The Ninth Data Release of the Sloan Digital Sky Survey: First Spectroscopic Data from the SDSS-III Baryon Oscillation Spectroscopic Survey
• Astrophysics Source Code Library
• The Palomar Transient Factory: System Overview, Performance and First Results
• The MAPPINGS III Library of Fast Radiative Shock Models
• Evolutionary Stellar Population Synthesis with MILES. Part I: The Base Models and a New Line Index System
• Spectral Classification and Redshift Measurement for the SDSS-III Baryon Oscillation Spectroscopic Survey
• The Sloan Lens ACS Survey. I. A Large Spectroscopically Selected Sample of Massive Early-Type Lens Galaxies
• Nearby Supernova Rates from the Lick Observatory Supernova Search. III. The Rate-Size Relation, and the Rates as a Function of Galaxy Hubble Type and Colour
• Kiloparsec-scale outflows are prevalent among luminous AGN: outflows and feedback in the context of the overall AGN population
• A Substantial Population of Low Mass Stars in Luminous Elliptical Galaxies
• ANNz: estimating photometric redshifts using artificial neural networks
• The Sloan Lens ACS Survey. V. The Full ACS Strong-Lens Sample
• SDSS DR7 White Dwarf Catalog
• Data Mining and Machine Learning in Astronomy
• Quasar feedback and the origin of radio emission in radio-quiet quasars
• The nature of LINER galaxies: Ubiquitous hot old stars and rare accreting black holes
• K-corrections and Extinction Corrections for Type Ia Supernovae
• Space density of optically-selected type 2 quasars
• Completion of a Survey and Detailed Study of Double-Peaked Emission Lines in Radio-Loud AGNs
• Ionised outflows in z $\sim$ 2.4 quasar host galaxies
• Aperture effects on Star Formation Rate, Metallicity and Reddening
• Double-Peaked Low-Ionization Emission Lines in Active Galactic Nuclei
• Feedback in Luminous Obscured Quasars
• Discovery of extreme [OIII]5007A outflows in high-redshift red quasars
• Accretion and star formation rates in low redshift type-II active galactic nuclei
• Revisiting the Hubble sequence in the SDSS DR7 spectroscopic sample: a publicly available bayesian automated classification
• A Search for Binary Active Galactic Nuclei: Double-Peaked [OIII] AGN in the Sloan Digital Sky Survey
• Suppressing star formation in quiescent galaxies with supermassive black hole winds
• Photometric Supernova Classification With Machine Learning
• The BOSS Emission-Line Lens Survey (BELLS). I. A large spectroscopically selected sample of Lens Galaxies at redshift ~ 0.5
• Photometric and Spectroscopic Properties of Type II-P Supernovae
• Type 2 AGNs with Double-Peaked [O III] Lines: Narrow Line Region Kinematics or Merging Supermassive Black Hole Pairs?
• Improved and Quality-assessed Emission and Absorption Line measurements in Sloan Digital Sky Survey galaxies
• Active Galactic Nuclei with Double-peaked Narrow Lines: Are They Dual AGNs?
• Discovery of superstrong, fading, iron line emission and double-peaked Balmer lines of the galaxy SDSSJ0952+2143 - the light echo of a huge flare
• Collisional heating as the origin of filament emission in galaxy clusters
• Automating Discovery and Classification of Transients and Variable Stars in the Synoptic Survey Era
• Automated Transient Identification in the Dark Energy Survey
• DA white dwarfs in SDSS DR7 and a search for infrared excess emission
• Spectral Decomposition of Broad-Line AGNs and Host Galaxies
• Hdelta-Selected Galaxies in the Sloan Digital Sky Survey I: The Catalog
• Extreme Coronal Line Emitters: Tidal Disruption of Stars by Massive Black Holes in Galactic Nuclei?
• The chemical diversity of exo-terrestrial planetary debris around white dwarfs
• Principal Component Analysis with Noisy and/or Missing Data
• Discovery of 90 Type Ia supernovae among 700,000 Sloan spectra: the Type-Ia supernova rate versus galaxy mass and star-formation rate at redshift ~0.1
• A Catalog of Local E+A(post-starburst) Galaxies selected from the Sloan Digital Sky Survey Data Release 5
• Double-peaked narrow emission-line galaxies from Sloan Digital Sky Survey. I. Sample and basic properties
• Sloan Digital Sky Survey Spectroscopic Lens Search. I. Discovery of Intermediate-Redshift Star-Forming Galaxies Behind Foreground Luminous Red Galaxies
• Estimation of stellar atmospheric parameters from SDSS/SEGUE spectra
• Systematic search for extremely metal poor galaxies in the Sloan Digital Sky Survey
• Shocked POststarbust Galaxy Survey I: Candidate Poststarbust Galaxies with Emission Line Ratios Consistent with Shocks
• Narrow double-peaked emission lines of SDSS J131642.90+175332.5: signature of a single or a binary AGN in a merger, jet-cloud interaction, or unusual narrow-line region geometry
• Transient Super-strong Coronal Lines and Broad Bumps in the Galaxy SDSS J074820.67+471214.3
• Automatic unsupervised classification of all SDSS/DR7 galaxy spectra
• Ultracompact AM CVn Binaries from the Sloan Digital Sky Survey: Three Candidates Plus the First Confirmed Eclipsing System
• The BOSS Emission-Line Lens Survey. III. : Strong Lensing of Ly$\alpha$ Emitters by Individual Galaxies
• A unified explanation for the supernova rate-galaxy mass dependency based on supernovae discovered in Sloan galaxy spectra
• Machine learning for transient discovery in Pan-STARRS1 difference imaging
• Spectroscopic detection of Type Ia Supernovae in the Sloan Digital Sky Survey
• Spectroscopic Identification of Type 2 Quasars at Z < 1 in SDSS-III/BOSS
• Black hole mass and accretion rate of active galactic nuclei with double-peaked broad emission lines
• A sample of Seyfert-2 galaxies with ultra-luminous galaxy-wide NLRs -- Quasar light echos?
• Are E+A Galaxies Dusty-Starbursts?: VLA 20cm Radio Continuum Observation
• Reducing the Dimensionality of Data: Locally Linear Embedding of Sloan Galaxy Spectra
• Near-infrared imaging of 222 nearby Hdelta-strong galaxies from the SDSS
• AGN with strong forbidden high-ionisation lines selected from the Sloan Digital Sky Survey
• Semi-supervised Learning for Photometric Supernova Classification
• Automated Classification of Periodic Variable Stars detected by the Wide-field Infrared Survey Explorer
• On the nature of sodium excess objects. I. Data and observed trends
• Unsupervised clustering of Type II supernova light curves
• SDSSJ140228.22+632133.3: A New Spectroscopically Selected Gravitational Lens
• Exploiting Low-Dimensional Structure in Astronomical Spectra
• Classification of Stellar Spectra with LLE
• A data-driven model for spectra: Finding double redshifts in the Sloan Digital Sky Survey
• A High Resolution Atlas of Composite SDSS Galaxy Spectra
• Automated unsupervised classification of the Sloan Digital Sky Survey stellar spectra using k-means clustering
• Clustering Based Feature Learning on Variable Stars
• IR Mergers and IR QSOs with Galactic Winds. II. NGC 5514, two Extra-nuclear Starburts/LINERs with a Supergiant Bubble in the Rupture Phase
• Multivariate Classification with Random Forests for Gravitational Wave Searches of Black Hole Binary Coalescence
• The Data Big Bang and the Expanding Digital Universe: High-Dimensional, Complex and Massive Data Sets in an Inflationary Epoch
• Using machine learning to classify the diffuse interstellar bands
• The spectral energy distributions of K+A galaxies from the UV to the mid-IR: stellar populations, star formation and hot dust
• Do galaxies form a spectroscopic sequence?
• Active Galactic Nuclei with Double-Peaked Balmer Lines: I. Black Hole Masses and the Eddington ratios
• Spectroscopic Determination of the Low Redshift Type Ia Supernova Rate from the Sloan Digital Sky Survey
• SDSS galaxies with double-peaked emission lines: double starbursts or AGNs?
• A Chandra Look at Five of the Broadest Double-Peaked Balmer-Line Emitters
• Soft clustering analysis of galaxy morphologies: A worked example with SDSS
• The Synthetic-Oversampling Method: Using Photometric Colors to Discover Extremely Metal-Poor Stars

Cited by in corpus (86)

• Probabilistic Random Forest: A machine learning algorithm for noisy datasets
• The Automatic Learning for the Rapid Classification of Events (ALeRCE) Alert Broker
• Surveying the reach and maturity of machine learning and artificial intelligence in astronomy
• The DAWES review 10: The impact of deep learning for the analysis of galaxy surveys
• An automatic taxonomy of galaxy morphology using unsupervised machine learning
• Radio Galaxy Zoo: Compact and extended radio source classification with deep learning
• Dimensionality Reduction of SDSS Spectra with Variational Autoencoders
• Astronomaly: Personalised Active Anomaly Detection in Astronomical Data
• Quasar Feedback Survey: Multi-phase outflows, turbulence and evidence for feedback caused by low power radio jets inclined into the galaxy disk
• Evolution Through the Post-Starburst Phase: Using Post-Starburst Galaxies as Laboratories for Understanding the Processes that Drive Galaxy Evolution
• Extragalactic Radio Continuum Surveys and the Transformation of Radio Astronomy
• Practical Galaxy Morphology Tools from Deep Supervised Representation Learning
• Direct evidence of AGN-feedback: a post starburst galaxy stripped of its gas by AGN-driven winds
• Detecting outliers in astronomical images with deep generative networks
• Anomaly detection in the Zwicky Transient Facility DR3
• Black hole mass estimation for Active Galactic Nuclei from a new angle
• Discovering the Unexpected in Astronomical Survey Data
• Systematic Serendipity: A Test of Unsupervised Machine Learning as a Method for Anomaly Detection
• Active Anomaly Detection for time-domain discoveries
• The use of convolutional neural networks for modelling large optically-selected strong galaxy-lens samples
• A review of unsupervised learning in astronomy
• Unveiling the rarest morphologies of the LOFAR Two-metre Sky Survey radio source population with self-organised maps
• Detecting outliers and learning complex structures with large spectroscopic surveys - a case study with APOGEE stars
• Anomaly Detection in the Open Supernova Catalog
• HOLISMOKES -- IV. Efficient mass modeling of strong lenses through deep learning
• Multi-phase outflows in post starburst E+A galaxies -- I. General sample properties and the prevalence of obscured starbursts
• A probabilistic approach to emission-line galaxy classification
• Multi-phase outflows in post starburst E+A galaxies -- II. A direct connection between the neutral and ionized outflow phases
• Evidence of ongoing AGN-driven feedback in a quiescent post starburst E+A galaxy
• Anomaly detection in Hyper Suprime-Cam galaxy images with generative adversarial networks
• Automated novelty detection in the WISE survey with one-class support vector machines
• Photometric redshifts with machine learning, lights and shadows on a complex data science use case
• Quasar Lenses and Galactic Streams: Outlier Selection and GAIA Multiplet Detection
• A method for finding anomalous astronomical light curves and their analogs
• Autoencoding Galaxy Spectra II: Redshift Invariance and Outlier Detection
• PSFGAN: a generative adversarial network system for separating quasar point sources and host galaxy light
• A semi-supervised Machine Learning search for never-seen Gravitational-Wave sources
• A Robust Study of High-Redshift Galaxies: Unsupervised Machine Learning for Characterising morphology with JWST up to z ~ 8
• 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
• Preparing to discover the unknown with Rubin LSST -- I: Time domain
• Neural network-based anomaly detection for high-resolution X-ray spectroscopy
• Rapid sorting of radio galaxy morphology using Haralick features
• Unsupervised classification of SDSS galaxy spectra
• Eigengalaxies: describing galaxy morphology using principal components in image space
• Astronomaly at scale: searching for anomalies amongst 4 million galaxies
• Machines Learn to Infer Stellar Parameters Just by Looking at a Large Number of Spectra
• HOLISMOKES -- IX. Neural network inference of strong-lens parameters and uncertainties from ground-based images
• Density Based Outlier Scoring on Kepler Data
• Stellar Spectral Interpolation using Machine Learning
• Identifying complex sources in large astronomical data using a coarse-grained complexity measure
• Exploring X-ray variability with unsupervised machine learning I. Self-organizing maps applied to XMM-Newton data
• Identifying outliers in astronomical images with unsupervised machine learning
• Searching for Anomalies in the ZTF Catalog of Periodic Variable Stars
• Identifying anomalous radio sources in the EMU Pilot Survey using a complexity-based approach
• Not Hydro: Using Neural Networks to estimate galaxy properties on a Dark-Matter-Only simulation
• ULISSE: A Tool for One-shot Sky Exploration and its Application to Active Galactic Nuclei Detection
• Deploying Jupyter Notebooks at scale on XSEDE resources for Science Gateways and workshops
• Redshifted broad absorption line quasars found via machine-learned spectral similarity
• NuSTAR hard X-ray data and Gemini 3D spectra reveal powerful AGN and outflow histories in two low-redshift Lyman-$\alpha$ blobs
• Dimension Estimation Using Autoencoders
• Galaxy Spectra Networks (GaSNet). III. Generative pre-trained network for spectrum reconstruction, redshift estimate and anomaly detection
• Predicting bulge to total luminosity ratio of galaxies using deep learning
• Characterization of Supernovae Based on the Spectral-Temporal Energy Distribution: Possible two SN Ib Subtypes
• The emptiness inside: Finding gaps, valleys, and lacunae with geometric data analysis
• Anomaly detection in Astrophysics: a comparison between unsupervised Deep and Machine Learning on KiDS data
• Comparison of outlier detection methods on astronomical image data
• One of Everything: The Breakthrough Listen Exotica Catalog
• Spectroscopic observations of the machine-learning selected anomaly catalogue from the AllWISE Sky Survey
• Disentangling stellar atmospheric parameters in astronomical spectra using Generative Adversarial Neural Networks
• A statistical study of lopsided galaxies using random forest
• Astroinformatics Challenges from Next-generation Radio Continuum Surveys
• Machine Learning Classification of Young Stellar Objects and Evolved Stars in the Magellanic Clouds Using the Probabilistic Random Forest Classifier
• Interpreting anomaly detection of SDSS spectra
• Launching the VASCO citizen science project
• Exploring the Universe with SNAD: Anomaly Detection in Astronomy
• ConvoSource: Radio-Astronomical Source-Finding with Convolutional Neural Networks
• Identification of Anomalous E+A Galaxies in GAMA Using an Isolation Forest
• ULISSE: Determination of star-formation rate and stellar mass based on the one-shot galaxy imaging technique
• The PAU Survey: Classifying low-z SEDs using Machine Learning clustering
• A Treasure Hunt in the Pool of Spectra Classified as High-redshift QSOs by the Spectroscopic Pipeline of the SDSS DR16
• Identifying Anomalous DESI Galaxy Spectra with a Variational Autoencoder
• Searching for outliers in the Chandra Source Catalog
• Where's Swimmy?: Mining unique color features buried in galaxies by deep anomaly detection using Subaru Hyper Suprime-Cam data
• J-PLUS: Understanding outlier white dwarfs in the third data release via dimensionality reduction
• Accurate spectroscopic redshift estimation using non-negative matrix factorization: application to MUSE spectra