¨My current research is focused on the Astrostatistics, applied to the fields of distance scale using Cepheids as indicators and classification of light curves of variable stars. In the first field, I am working on the non-linearity of the period-luminosity relation for Cepheids. By using robust linear regressions, it is possible to find that this relation is linear.
Currently, large astronomical databases are beeing generated, powered by the development wide field of view detectors. The area of variable stars is a good example of that, and it is necessary to have automatic classifcation techniques of stars. One of the best methods is the supervised classification, such as random forests, K-Nearest Neighbors, vector support machines, among others. Finding the special features that optimize these methods allows to obtain the most efficient and faster mechanisms to classify and select variable stars. This is the aim of my reserach on this topic, as well as applying these methods to classify variable stars like Cepheids, Be stars, eclipsing binaries, long period variables, among others. This is an interdisciplinary investigation, where I collaborate with researchers and students of Industrial Engineering, Mathematics and Statistics.
Other fields of my interest in which I am doing or have done investigation are astrometry, distance scale and stellar clusters¨.