A Hybrid Ensemble Learning Approach to Star-Galaxy Classification

Abstract

There exist a variety of star-galaxy classification techniques, each with their own strengths and weaknesses. In this paper, we present a novel meta-classification framework that combines and fully exploits different techniques to produce a more robust star-galaxy classification. To demonstrate this hybrid, ensemble approach, we combine a purely morphological classifier, a supervised machine learning method based on random forest, an unsupervised machine learning method based on self-organizing maps, and a hierarchical Bayesian template fitting method. Using data from the CFHTLenS survey, we consider different scenarios: when a high-quality training set is available with spectroscopic labels from DEEP2, SDSS, VIPERS, and VVDS, and when the demographics of sources in a low-quality training set do not match the demographics of objects in the test data set. We demonstrate that our Bayesian combination technique improves the overall performance over any individual classification method in these scenarios. Thus, strategies that combine the predictions of different classifiers may prove to be optimal in currently ongoing and forthcoming photometric surveys, such as the Dark Energy Survey and the Large Synoptic Survey Telescope.

arXiv comment15 pages, 18 figures. Accepted for publication in MNRAS. Code available at https://github.com/EdwardJKim/astroclass

References in corpus (27)

• emcee: The MCMC Hammer
• HEALPix -- a Framework for High Resolution Discretization, and Fast Analysis of Data Distributed on the Sphere
• Evolutionary models for very-low-mass stars and brown dwarfs with dusty atmospheres
• Bayesian photometric redshift estimation
• The Tenth Data Release of the Sloan Digital Sky Survey: First Spectroscopic Data from the SDSS-III Apache Point Observatory Galactic Evolution Experiment
• A Method for Weak Lensing Observations
• The DEEP2 Galaxy Redshift Survey: Design, Observations, Data Reduction, and Redshifts
• CFHTLenS: The Canada-France-Hawaii Telescope Lensing Survey
• Issues in Stacked Generalization
• The VIMOS VLT Deep Survey - First epoch VVDS-Deep survey: 11564 spectra with 17.5<=IAB<=24, and the redshift distribution over 0< z <=5
• CFHTLenS: The Canada-France-Hawaii Telescope Lensing Survey - Imaging Data and Catalogue Products
• CFHTLenS: Improving the quality of photometric redshifts with precision photometry
• TPZ : Photometric redshift PDFs and ancillary information by using prediction trees and random forests
• The VIMOS Public Extragalactic Survey (VIPERS): First Data Release of 57 204 spectroscopic measurements
• Ameliorating Systematic Uncertainties in the Angular Clustering of Galaxies: A Study using SDSS-III
• The CFHT Legacy Survey: stacked images and catalogs
• The Vimos VLT Deep Survey: Global properties of 20000 galaxies in the I_AB<=22.5 WIDE survey
• SOMz: photometric redshift PDFs with self organizing maps and random atlas
• Robust Machine Learning Applied to Astronomical Datasets I: Star-Galaxy Classification of the SDSS DR3 Using Decision Trees
• Decision Tree Classifiers for Star/Galaxy Separation
• Star/galaxy separation at faint magnitudes: Application to a simulated Dark Energy Survey
• Star-Galaxy Classification in Multi-Band Optical Imaging
• Exhausting the Information: Novel Bayesian Combination of Photometric Redshift PDFs
• A Census of Object Types and Redshift Estimates in the SDSS Photometric Catalog from a Trained Decision-Tree Classifier
• A Bayesian approach to star-galaxy classification
• The stellar content of the COSMOS field as derived from morphological and SED based gtar/galaxy Separation
• Effect of training characteristics on object classification: an application using Boosted Decision Trees

Cited by in corpus (33)

• The Dark Energy Survey Data Release 1
• Star-galaxy Classification Using Deep Convolutional Neural Networks
• Probabilistic Random Forest: A machine learning algorithm for noisy datasets
• Deblending and Classifying Astronomical Sources with Mask R-CNN Deep Learning
• PS1-STRM: Neural network source classification and photometric redshift catalogue for PS1 $3\pi$ DR1
• Machine Learning and Cosmological Simulations II: Hydrodynamical Simulations
• The miniJPAS survey: star-galaxy classification using machine learning
• J-PLUS: Morphological star/galaxy classification by PDF analysis
• Machine Learning and Cosmological Simulations I: Semi-Analytical Models
• A review of unsupervised learning in astronomy
• A cooperative approach among methods for photometric redshifts estimation: an application to KiDS data
• The Dark Energy Survey 5-year photometrically identified Type Ia Supernovae
• Star-galaxy classification in the Dark Energy Survey Y1 dataset
• Deep Horizon; a machine learning network that recovers accreting black hole parameters
• COSMOS2020: Manifold Learning to Estimate Physical Parameters in Large Galaxy Surveys
• The PAU Survey: star-galaxy classification with multi narrow-band data
• Stacking for machine learning redshifts applied to SDSS galaxies
• The Photometric LSST Astronomical Time-series Classification Challenge (PLAsTiCC): Selection of a performance metric for classification probabilities balancing diverse science goals
• Exploring Photometric Redshifts as an Optimization Problem: An Ensemble MCMC and Simulated Annealing-Driven Template-Fitting Approach
• Inferring physical properties of galaxies from their emission line spectra
• The High Cadence Transient Survey (HITS): Compilation and characterization of light-curve catalogs
• SKA Science Data Challenge 2: analysis and results
• Candidate Massive Galaxies at $z\sim 4$ in the Dark Energy Survey
• Star-Galaxy Image Separation with Computationally Efficient Gaussian Process Classification
• Classifying Stars, Galaxies and AGN in CLAUDS+HSC-SSP Using Gradient Boosted Decision Trees
• Morphological Star-Galaxy Separation
• Estimation of Photometric Redshifts. I. Machine Learning Inference for Pan-STARRS1 Galaxies Using Neural Networks
• Reliable Photometric Membership (RPM) of Galaxies in Clusters. I. A Machine Learning Method and its Performance in the Local Universe
• The miniJPAS survey quasar selection III: Classification with artificial neural networks and hybridisation
• Preparation for CSST: Star-galaxy Classification using a Rotationally Invariant Supervised Machine Learning Method
• Detecting Hyperons in neutron stars -- a machine learning approach
• Star-Galaxy Separation via Gaussian Processes with Model Reduction
• Unsupervised Star Galaxy Classification with Cascade Variational Auto-Encoder