Three-Dimensional Spectral Classification of Low-Metallicity Stars Using Artificial Neural Networks

Abstract

We explore the application of artificial neural networks (ANNs) for the estimation of atmospheric parameters (Teff, logg, and [Fe/H]) for Galactic F- and G-type stars. The ANNs are fed with medium-resolution (~ 1-2 A) non flux-calibrated spectroscopic observations. From a sample of 279 stars with previous high-resolution determinations of metallicity, and a set of (external) estimates of temperature and surface gravity, our ANNs are able to predict Teff with an accuracy of ~ 135-150 K over the range 4250 <= Teff <= 6500 K, logg with an accuracy of ~ 0.25-0.30 dex over the range 1.0 <= logg <= 5.0 dex, and [Fe/H] with an accuracy ~ 0.15-0.20 dex over the range -4.0 <= [Fe/H] <= +0.3. Such accuracies are competitive with the results obtained by fine analysis of high-resolution spectra. It is noteworthy that the ANNs are able to obtain these results without consideration of photometric information for these stars. We have also explored the impact of the signal-to-noise ratio (S/N) on the behavior of ANNs, and conclude that, when analyzed with ANNs trained on spectra of commensurate S/N, it is possible to extract physical parameter estimates of similar accuracy with stellar spectra having S/N as low as 13. Taken together, these results indicate that the ANN approach should be of primary importance for use in present and future large-scale spectroscopic surveys.

51 pages, 11 eps figures, uses aastex; to appear in ApJ

References in corpus (2)

• GAIA: Composition, Formation and Evolution of the Galaxy
• A database of high and medium-resolution stellar spectra

Cited by in corpus (26)

• The Milky Way Tomography with SDSS: II. Stellar Metallicity
• The Hamburg/ESO R-process Enhanced Star survey (HERES) II. Spectroscopic analysis of the survey sample
• ULySS: A Full Spectrum Fitting Package
• Coude-feed stellar spectral library - atmospheric parameters
• Automated derivation of stellar atmospheric parameters and chemical abundances: the MATISSE algorithm
• Metal Abundances and Kinematics of Bright Metal-Poor Giants Selected from the LSE Survey: Implications for the Metal-Weak Thick Disk
• Estimation of stellar atmospheric parameters from SDSS/SEGUE spectra
• 3D NLTE spectral line formation of lithium in late-type stars
• Catalogue of new Herbig Ae/Be and classical Be stars. A machine learning approach to Gaia DR2
• The Isotopic Mixture of Barium in the Metal-poor Subgiant HD 140283
• Estimation of Carbon Abundances in Metal-Poor Stars. I. Application to the "Strong G-band" stars of Beers, Preston, & Shectman
• Statistical methods of automatic spectral classification and their application to the Hamburg/ESO Survey
• The ILIUM forward modelling algorithm for multivariate parameter estimation and its application to derive stellar parameters from Gaia spectrophotometry
• WIYN Open Cluster Study, LXXVI. Li Evolution Among Stars of Low/Intermediate Mass: The Metal-Deficient Open Cluster, NGC 2506
• Analysis of medium resolution spectra by automated methods - application to M55 and omega Centauri
• Stellar atmosphere parameters with MAx, a MAssive compression of x^2 for spectral fitting
• SDSS/SEGUE Spectral Feature Analysis For Stellar Atmospheric Parameter Estimation
• On the estimation of stellar parameters with uncertainty prediction from Generative Artificial Neural Networks: application to Gaia RVS simulated spectra
• Automatic classification of eclipsing binaries light curves using neural networks
• Transit Clairvoyance: Enhancing TESS follow-up using artificial neural networks
• An Automated System to Classify Stellar Spectra I
• Parameterizing Stellar Spectra Using Deep Neural Networks
• Neural network for determining an asteroid mineral composition from reflectance spectra
• Identification of metal-poor stars using the artificial neural network
• Automated Determination of Stellar Parameters from Simulated Dispersed Images for DIVA
• V474 Car: A Rare Halo RS CVn Binary in Retrograde Galactic Orbit