Stellar Parameters in an Instant with Machine Learning
Earl Bellinger (Max-Planck-Institut für Sonnensystemforschung), George Angelou (Max-Planck-Institut für Sonnensystemforschung), Saskia Hekker (Max-Planck-Institut für Sonnensystemforschung), Sarbani Basu (Yale University), et al.

With the advent of dedicated photometric space missions, the ability to rapidly process huge catalogues of stars has become paramount. We introduce a new method based on machine learning for inferring the stellar parameters of main-sequence stars exhibiting solar-like oscillations. Our method makes precise predictions that are competitive with other methods, but with the advantage of costing practically no time. We validate our technique on a hare-and-hound exercise, the Sun, and 16 Cygni and then use it to predict the parameters of the Kepler objects-of-interest. Finally, we present novel insights into main-sequence evolution that have been extracted by the algorithm.