An Application of Supervised Learning Methods to Search for Variable Stars in a Selected Field of the VVV Survey
Abstract
We characterize properties of time series of variable stars in the B278 field of the VVV survey, using robust statistics. Using random forest and support vector machines classifiers we propose 47 candidates to RR Lyraae, and 12 candidates to WU Ursae Majoris eclipsing binaries.
- Publication:
-
Revista Mexicana de Astronomia y Astrofisica Conference Series
- Pub Date:
- July 2017
- Bibcode:
- 2017RMxAC..49...97E
- Keywords:
-
- variables: general;
- techniques: miscellaneous