Maria Tsantaki
IA

Abstract
In the near future, large spectroscopic surveys (Gaia, 4MOST, WEAVE, MSE) will provide millions of stellar spectra with the goal to precisely characterize stars in terms of their atmospheric parameters (effective temperature, surface gravity and metallicity). However, to deal with the unprecedented amount of data, new analysis techniques need to be introduced since our standard methods are limited due to the required computational time.
In this talk, I will explain a method to derive stellar parameters based on the principles of machine learning. I have costumed this method to be applied to the wavelength range and resolution of the expected Gaia-RVS spectra (847-871 nm, R~11200). The training sample is constructed from a grid of synthetic spectra covering a large parameter space. I will present comparisons of the results of this method with real data and evaluate the goodness of this method.

2019 April 05, 13:30

IA/U.Porto
Centro de Astrofísica da Universidade do Porto (Classroom)
Rua das Estrelas, 4150-762 Porto