Automated Classification of Stellar Spectra. II: Two-Dimensional Classification with Neural Networks and Principal Components Analysis

Abstract

We investigate the application of neural networks to the automation of MK spectral classification. The data set for this project consists of a set of over 5000 optical (3800-5200 AA) spectra obtained from objective prism plates from the Michigan Spectral Survey. These spectra, along with their two-dimensional MK classifications listed in the Michigan Henry Draper Catalogue, were used to develop supervised neural network classifiers. We show that neural networks can give accurate spectral type classifications (sig_68 = 0.82 subtypes, sig_rms = 1.09 subtypes) across the full range of spectral types present in the data set (B2-M7). We show also that the networks yield correct luminosity classes for over 95% of both dwarfs and giants with a high degree of confidence. Stellar spectra generally contain a large amount of redundant information. We investigate the application of Principal Components Analysis (PCA) to the optimal compression of spectra. We show that PCA can compress the spectra by a factor of over 30 while retaining essentially all of the useful information in the data set. Furthermore, it is shown that this compression optimally removes noise and can be used to identify unusual spectra.

To appear in MNRAS. 15 pages, 17 figures, 7 tables. 2 large figures (nos. 4 and 15) are supplied as separate GIF files. The complete paper can be obtained as a single gziped PS file from http://wol.ra.phy.cam.ac.uk/calj/p1.html

Cited by in corpus (59)

• The Milky Way Tomography with SDSS: II. Stellar Metallicity
• A Unified Near Infrared Spectral Classification Scheme for T Dwarfs
• Data Mining and Machine Learning in Astronomy
• Estimating photometric redshifts with artificial neural networks
• Distributions of Galaxy Spectral Types in the Sloan Digital Sky Survey
• Galaxy Zoo: Reproducing Galaxy Morphologies Via Machine Learning
• Improving the Estimation of Star formation Rates and Stellar Population Ages of High-redshift Galaxies from Broadband Photometry
• Comparative performance of selected variability detection techniques in photometric time series
• Estimation of stellar atmospheric parameters from SDSS/SEGUE spectra
• PCA Tomography: how to extract information from datacubes
• Application of Convolutional Neural Networks for Stellar Spectral Classification
• Light curve analysis of Variable stars using Fourier decomposition and Principal component analysis
• The Galah Survey: Classification and diagnostics with t-SNE reduction of spectral information
• Neural Networks and Photometric Redshifts
• Reducing the Dimensionality of Data: Locally Linear Embedding of Sloan Galaxy Spectra
• Exploring the Morphology of RAVE Stellar Spectra
• Three-Dimensional Spectral Classification of Low-Metallicity Stars Using Artificial Neural Networks
• Properties of Dark Matter Haloes and their Correlations: the Lesson from Principal Component Analysis
• Automated Clustering Algorithms for Classification of Astronomical Objects
• A review of unsupervised learning in astronomy
• The INT Search for Metal-Poor Stars. Spectroscopic Observations and Classification via Artificial Neural Networks
• Statistical methods of automatic spectral classification and their application to the Hamburg/ESO Survey
• Principal Component Analysis of SDSS Stellar Spectra
• Spectral modeling of type II supernovae II. A machine learning approach to quantitative spectroscopic analysis
• Classification of Stellar Spectra with LLE
• A new method for the inversion of atmospheric parameters of A/Am stars
• Inversion of stellar fundamental parameters from Espadons and Narval high-resolution spectra
• Stellar Spectra Classification and Feature evaluation Based on Random Forest
• Powerful t-SNE technique leading to clear separation of type-2 AGN and HII galaxies in BPT diagrams
• An autoencoder of stellar spectra and its application in automatically estimating atmospheric parameters
• Classification and redshift estimation by principal component analysis
• Machine learning in APOGEE: Unsupervised spectral classification with $K$-means
• Use of neural networks for the identification of new z>=3.6 QSOs from FIRST-SDSS DR5
• SDSS-DR12 Bulk Stellar Spectral Classification: Artificial Neural Networks Approach
• A metric space for type Ia supernova spectra
• Rapid sorting of radio galaxy morphology using Haralick features
• Machine learning in APOGEE: Identification of stellar populations through chemical abundances
• Machines Learn to Infer Stellar Parameters Just by Looking at a Large Number of Spectra
• Eigenvector Sky Subtraction
• Deep Learning in Searching the Spectroscopic Redshift of Quasars
• Nonparametric Transient Classification using Adaptive Wavelets
• The Gaia-ESO Survey: Preparing the ground for 4MOST & WEAVE galactic surveys. Chemical evolution of lithium with machine learning
• A Spectral Classification System for Hydrogen-deficient Carbon Stars
• Stellar Spectral Interpolation using Machine Learning
• Random Forest classification of Gaia DR3 white dwarf-main sequence spectra: a feasibility study
• The GALAH Survey: Stellar parameters and abundances for 800,000 Gaia RVS spectra using GALAH DR4 and The Cannon
• Variability search in M 31 using Principal Component Analysis and the Hubble Source Catalog
• High-precision interpolation of stellar atmospheres with a deep neural network using a 1D convolutional auto encoder for feature extraction
• Selection of quasar candidates from combined radio and optical surveys using neural networks
• Data-Space Validation of High-Dimensional Models by Comparing Sample Quantiles
• Stellar Karaoke: deep blind separation of terrestrial atmospheric effects out of stellar spectra by velocity whitening
• Automated Determination of Stellar Parameters from Simulated Dispersed Images for DIVA
• Relativistic Reflection Modeling in AGN and Related Variability from PCA: A Brief Review
• Disentangling stellar atmospheric parameters in astronomical spectra using Generative Adversarial Neural Networks
• Harvesting the Lyα forest with convolutional neural networks
• Objective Subclass Determination of Sloan Digital Sky Survey Unknown Spectral Objects
• Physical Parameters of Stars in NGC 6397 Using ANN-Based Interpolation and Full Spectrum Fitting
• Astrometric Binary Classification Via Artificial Neural Networks
• Spectral Classification; Old and Contemporary