Description
The Magellanic Clouds are two nearby, gas-rich, and metal-poor galaxies, characterised by extended regions of star formation (SF). Thanks to their proximity, they represent ideal targets for the study of resolved star-formation processes in a metal-poor environment.
For this work we analysed near-infrared data obtained with the VISTA Survey of the Magellanic Clouds (VMC, Cioni et al. 2011) on a region of 1.5 deg$^2$ in the Large Magellanic Cloud (LMC), combined with more recent proprietary observations of the same region (LMC Tile 7_5), as well as optical photometry data from the SMASH survey (Nidever et al., 2017), resulting in a dataset with photometric depth and a wide wavelength baseline. Employing machine learning techniques we identified low mass pre main-sequence (PMS) stars to act as tracers of SF. We applied a probabilistic random forest algorithm (PRF, Reis et al., 2019) to classify sources into three target classes (PMS, upper main-sequence, and field stars) with an overall accuracy higher that 90$\%$. A clustering analysis allowed us to recover the most prominent SF structure of the region already known in literature (N44 and N51), while also identifying possible new SF sites. Isochrone fitting methods allowed us to determine the mass and ages of PMS stars in the SF regions, which, combined with their clustering properties, allowed us to characterise the temporal and spatial progression of SF across the tile region.
$\bullet$ Cioni, et al., 2011, A&A, 527, A116
$\bullet$ Nidever et al., 2017, AJ, 154:199
$\bullet$ Reis et al., 2019, AJ, 157, 16
