| A Deep Learning Neural Network Algorithm for the Classification of Eclipsing Binary Light Curves |
| Burak Ulaş |
| Konkoly Observatory/Çanakkale Onsekiz Mart University; Turkey |
| Machine learning techniques strengthen their solid ground in various areas from art to science every passing day. In the present, countless machine learning applications let scientists achieve faster and more precise results in many areas such as astrophysics where the researchers already made efforts to detect, fit, and classify the light curves of binary systems using machine and deep learning algorithms. In this talk, I present an image classification algorithm using deep learning convolutional neural network architecture, which classifies the morphologies of eclipsing binary systems based on their light curves. The algorithm trains the machine with light curve images generated from the observational data of eclipsing binary stars in contact, detached and semi-detached morphologies, whose light curves are provided by several catalogs. I give some details on the structure of the architecture and discuss the parameters of the network layers and the quality of the resulting metrics. I also plan to discourse upon an extended learning model to estimate the light curve parameters beside geometrical configuration. |