| Title: | Astronomical time-series and machine learning |
| Authors: | Kevin Guegan, Author |
| Material Type: | ISU Individual Project |
| Publisher: | Illkirch-Graffenstaden (France) : International Space University, 2021 |
| Format: | 1 electronic resource / col. ill. |
| Bibliography note: | Includes bibliographical references |
| Languages: | English |
| Description: | In this rapport is presented a search on variable stars (star whose the luminosity varies) thank to an algorithm of Machine Learning: UPSILoN. This algorithm works by the extraction of 16 different features defined from the light curve of the stars to give an automated classification followed by a percentage of probability. The project has been performed over a dataset composed of 167 520 stars from the 17th quarter of the space telescope Kepler. The interest of those specific data is that they have not been collected for the search of pulsating or rotational variable stars but for exoplanet, giant black hole or exploded stars. Inside of this dataset, a sample of 28 240 stars has been identified in which potential good candidates could be either supplement the knowledges that astrophysicists have through bibliography or be some completely new unknown variable stars. This rapport presents exactly the detail of 22 stars in which a method of confirmation detection has been applied, leading to 7 candidates for being new variable stars. Furthermore, as a sub research, a study has been led over 1850 Eclipsing Binary to identify exoplanets, from which appeared three candidates for being new exoplanets. Those 28 240 stars require all more particular research and are composed of Cepheid, RRLyrae variable stars and in majority of Eclipsing Binary and delta dcuti. The Long variable Period stars present in the dataset are not treated in this project. |
| ISU program : | Master of Space Studies |
| Format : | Online |
| Permalink: | https://isulibrary.isunet.edu/index.php?lvl=notice_display&id=11372 |
Read online (1)
 Guegan, Kevin_IP (1.33MB) Adobe Acrobat PDF |
