Cycle-StarNet: Bridging the gap between theory and data by leveraging large datasets

Abstract

The advancements in stellar spectroscopy data acquisition have made it necessary to accomplish similar improvements in efficient data analysis techniques. Current automated methods for analyzing spectra are either (a) data-driven, which requires prior knowledge of stellar parameters and elemental abundances, or (b) based on theoretical synthetic models that are susceptible to the gap between theory and practice. In this study, we present a hybrid generative domain adaptation method that turns simulated stellar spectra into realistic spectra by applying unsupervised learning to large spectroscopic surveys. We apply our technique to the APOGEE H-band spectra at R=22,500 and the Kurucz synthetic models. As a proof of concept, two case studies are presented. The first of which is the calibration of synthetic data to become consistent with observations. To accomplish this, synthetic models are morphed into spectra that resemble observations, thereby reducing the gap between theory and observations. Fitting the observed spectra shows an improved average reduced ${\chi }_R^2$ from 1.97 to 1.22, along with a reduced mean residual from 0.16 to -0.01 in normalized flux. The second case study is the identification of the elemental source of missing spectral lines in the synthetic modelling. A mock dataset is used to show that absorption lines can be recovered when they are absent in one of the domains. This method can be applied to other fields, which use large data sets and are currently limited by modelling accuracy. The code used in this study is made publicly available on github.

arXiv comment23 pages, 15 figures, 2 tables, accepted for publication on Nov 12, 2020, Nov 12. A companion 4-page preview is accepted to the ICML 2020 Machine Learning Interpretability for Scientific Discovery workshop. The code used in this study is made publicly available on github: https://github.com/teaghan/Cycle_SN

References in corpus (22)

• SEGUE: A Spectroscopic Survey of 240,000 stars with g=14-20
• The Cannon: A data-driven approach to stellar label determination
• The GALAH Survey: Second Data Release
• APOGEE Data Releases 13 and 14: Data and Analysis
• The Payne: self-consistent ab initio fitting of stellar spectra
• Abundance Estimates for 16 Elements in 6 Million Stars from LAMOST DR5 Low-Resolution Spectra
• Deep learning of multi-element abundances from high-resolution spectroscopic data
• High-Resolution Spectroscopy of Extremely Metal-Poor Stars from SDSS/SEGUE: I. Atmospheric Parameters and Chemical Compositions
• The atmospheric parameters and spectral interpolator for the stars of MILES
• The SDSS-III APOGEE Spectral Line List for H-band Spectroscopy
• Non-LTE oxygen line formation in 3D hydrodynamic model stellar atmospheres
• An Application of Deep Neural Networks in the Analysis of Stellar Spectra
• Project overview and update on WEAVE: the next generation wide-field spectroscopy facility for the William Herschel Telescope
• NLTE Chemical abundances in Galactic open and globular clusters
• Pristine IX: CFHT ESPaDOnS Spectroscopic Analysis of 115 Bright Metal-Poor Candidate Stars
• New atmospheric parameters and spectral interpolator for the MILES cool stars
• Measuring 14 elemental abundances with R=1,800 LAMOST spectra
• To use or not to use synthetic stellar spectra in population synthesis models?
• The RAdial Velocity Experiment (RAVE): Parameterisation of RAVE spectra based on convolutional neural networks
• Measuring oxygen abundances from stellar spectra without oxygen lines
• On line contribution functions and examining spectral line formation in 3D model stellar atmospheres
• The Peculiar Globular Cluster Palomar 1 and Persistence in the SDSS-APOGEE Database

Cited by in corpus (18)

• How Many Elements Matter?
• The Spectroscopic Binaries from LAMOST Medium-Resolution Survey (MRS). I. Searching for Double-lined Spectroscopic Binaries (SB2s) with Convolutional Neural Network
• A variational encoder-decoder approach to precise spectroscopic age estimation for large Galactic surveys
• Stellar Parameters and Chemical Abundances Estimated from LAMOST-II DR8 MRS based on Cycle-StarNet
• Quasar Factor Analysis -- An Unsupervised and Probabilistic Quasar Continuum Prediction Algorithm with Latent Factor Analysis
• Disentangled Representation Learning for Astronomical Chemical Tagging
• The Gaia-ESO Survey: Preparing the ground for 4MOST & WEAVE galactic surveys. Chemical evolution of lithium with machine learning
• Homogeneous Stellar Atmospheric Parameters and 22 Elemental Abundances for FGK Stars Derived From LAMOST Low-resolution Spectra with DD-Payne
• Closing the stellar labels gap: Stellar label independent evidence for [$\alpha /M$] information in Gaia BP/RP spectra
• Reliable stellar abundances of individual stars with the MUSE integral-field spectrograph
• Inferring stellar parameters and their uncertainties from high-resolution spectroscopy using invertible neural networks
• A Machine Learning Approach to Integral Field Unit Spectroscopy Observations: III. Disentangling Multiple Components in Hii regions
• PhDLspec: physical-prior embedded deep learning method for spectroscopic determination of stellar labels in high-dimensional parameter space
• PISP: Projected-Space Inference of Stellar Parameters
• Functional Data Analysis for Extracting the Intrinsic Dimensionality of Spectra: Application to Chemical Homogeneity in the Open Cluster M67
• Zeta-Payne: a fully automated spectrum analysis algorithm for the Milky Way Mapper program of the SDSS-V survey
• Can AI Dream of Unseen Galaxies? Conditional Diffusion Model for Galaxy Morphology Augmentation
• Playing CHESS with stars. I. Search for similar stars in large spectroscopic data sets