Poster abstract details

Unsupervised Machine Learning on the Selected Near Infrared Color Magnitude Diagrams of the Low Latitude Southern Galactic Plane
Efsan Sokmen, Iain Murray, Sebastián L. Hidalgo

Abstract

Most of the stars in our galaxy is contained within the disk, of which the maximum size and the detailed structure is still unknown. Using the near infrared bands sheds light on the complex stellar distribution of the galactic thin disk since it is less affected by the scatter and obscureness due to the dust. Using The ESO near-infrared public survey VISTA variables in the Vía Lactea (VVV) from which we had obtained the deep photometric data for the disk in J and $\mathrm{K_{s}}$ filters reaching 19 mag in $\mathrm{K_{s}}$, we apply the principal component analysis (PCA) to the color magnitude diagrams (CMDs) and investigate the significance of these results by comparing them to those that are applied on a set of synthetic populations with extinction. We found that the principal components correlate with extinction on the small scale, while on the large scale it can be used to find anomalies.