Deep Learning Applied to the Asteroseismic Modeling of Stars with Coherent Oscillation Modes

Abstract

We develop a novel method based on machine learning principles to achieve optimal initiation of CPU-intensive computations for forward asteroseismic modeling in a multi-D parameter space. A deep neural network is trained on a precomputed asteroseismology grid containing about 62 million coherent oscillation-mode frequencies derived from stellar evolution models. These models are representative of the core-hydrogen burning stage of intermediate-mass and high-mass stars. The evolution models constitute a 6D parameter space and their predicted low-degree pressure-and gravity-mode oscillations are scanned, using a genetic algorithm. A software pipeline is created to find the best fitting stellar parameters for a given set of observed oscillation frequencies. The proposed method finds the optimal regions in the 6D parameters space in less than a minute, hence providing the optimal starting point for further and more detailed forward asteroseismic modeling in a high-dimensional context. We test and apply the method to seven pulsating stars that were previously modeled asteroseismically by classical grid-based forward modeling based on a χ 2 statistic and obtain good agreement with past results. Our deep learning methodology opens up the application of asteroseismic modeling in +6D parameter space for thousands of stars pulsating in coherent modes with long lifetimes observed by the Kepler space telescope and to be discovered with the TESS and PLATO space missions, while applications so far were done star-by-star for only a handful of cases. Our method is open source and can be used by anyone freely a)

Asteroseismology of High-Mass Stars: New Insights of Stellar Interiors With Space Telescopes

“…Optimum models were selected using a genetic algorithm making such an automated pipeline extremely quick compared to more conventional forward seismic modeling techniques. Hendriks and Aerts (2019) test their methodology using several well-studied massive stars to benchmark the accuracy of their technique and good agreement is found overall. However, as discussed in detail by Hendriks and Aerts (2019), the modeling results based on their deep neural network depend on the choice of hyperparameters, which affect the ability to find the global minimum in the solution space.…”

“…Hendriks and Aerts (2019) test their methodology using several well-studied massive stars to benchmark the accuracy of their technique and good agreement is found overall. However, as discussed in detail by Hendriks and Aerts (2019), the modeling results based on their deep neural network depend on the choice of hyperparameters, which affect the ability to find the global minimum in the solution space. In Table 1 the model parameters resulting from the optimized tuning of these hyperparameters are provided (cf.…”

“…Recently, a new method of performing forward seismic modeling by means of a machine learning was developed by Hendriks and Aerts (2019). The authors trained a deep neural network using more than 62 million pulsation mode frequencies, which were calculated from a vast grid of stellar models covering main sequence intermediate-and high-mass stars spanning from 2 to 20 M ⊙ .…”

“…gray points), but such values are claimed to be an "optimal starting point" for further complex modeling. We refer the reader to Hendriks and Aerts (2019) for full details.…”

Hierarchically modelling Kepler dwarfs and subgiants to improve inference of stellar properties with asteroseismology

“…In this work, we train an artificial neural network (ANN) on a large grid of stellar models. There have been similar applications of ANNs in asteroseismology (Verma et al 2016;Bellinger et al 2016;Hon et al 2017Hon et al , 2018Hendriks & Aerts 2019) but not yet in the context of an HBM. Using the machine learning speed-up, we demonstrate a scalable method for obtaining fundamental stellar parameters.…”