Paper:	Stellar Parameters with Deep Learning
Volume:	522, Astronomical Data Analysis Software and Systems XXVII
Page:	393
Authors:	Fabbro, S.; Venn, K.; O'Briain, T.; Bialek, S.; Kielty, C.; Jahandar, F.; Monty, S.
Abstract:	Spectroscopic surveys require fast and efficient data analysis methods to maximize their scientific impact. Here we examine the predictive capabilities of deep neural network architectures applied on stellar spectra to derive physical parameters. We show how we train either on synthetic stellar spectra, on real data, and how we train on synthetic data and predict on real data to still achieve good accuracy stellar parameters within a wide range of signal to noise. We make extensive comparison to other methods with the SDSS APOGEE survey spectra. We also show how we are able to estimate uncertainties and automatically tune the neural network architectures.

Astronomical Society of the Pacific
Conference Series

© Copyright 1988 - 2024 - Astronomical Society of the Pacific
No part of the material protected by this copyright may be reproduced or used in any form other than for personal use without written permission from the ASP.

Page generated in 0.112 seconds