Machine Learning Classification of Young Stellar Objects and Evolved Stars in the Magellanic Clouds Using the Probabilistic Random Forest Classifier

The Magellanic Clouds (MCs) are excellent locations to study stellar dust emission and its contribution to galaxy evolution. Through spectral and photometric classification, MCs can serve as a unique environment for studying stellar evolution and galaxies enriched by dusty stellar point sources. We applied machine learning classifiers to spectroscopically labeled data from the Surveying the Agents of Galaxy Evolution (SAGE) project, which involved 12 multiwavelength filters and 618 stellar objects at the MCs. We classified stars into five categories: young stellar objects (YSOs), carbon-rich asymptotic giant branch (CAGB) stars, oxygen-rich AGB (OAGB) stars, red supergiants (RSG), and post-AGB (PAGB) stars. Following this, we augmented the distribution of imbalanced classes using the Synthetic Minority Oversampling Technique (SMOTE). Therefore, the Probabilistic Random Forest (PRF) classifier achieved the highest overall accuracy, reaching 89% based on the recall metric, in categorizing dusty stellar sources before and after data augmentation. In this study, SMOTE did not impact the classification accuracy for the CAGB, PAGB, and RSG categories but led to changes in the performance of the OAGB and YSO classes.