ML-MOC: Machine Learning (kNN and GMM) based Membership determination for Open Clusters

ABSTRACT The existing open-cluster membership determination algorithms are either prior dependent on some known parameters of clusters or not automatable to large samples clusters. In this paper, we present ml-moc, a new machine-learning-based approach identify likely members open using the Gaia DR2 data and no priori information about cluster parameters. We use k-nearest neighbour (kNN) algorithm Gaussian mixture model (GMM) high-precision proper motions parallax measurements from determine probabilities individual sources down G ? 20 mag. To validate developed method, apply it 15 clusters: M67, NGC 2099, 2141, 2243, 2539, 6253, 6405, 6791, 7044, 7142, 752, Blanco 1, Berkeley 18, IC 4651, Hyades. These differ in terms their ages, distances, metallicities, extinctions cover wide parameter space parallaxes with respect field population. extracted produce clean colour–magnitude diagrams our astrometric good agreement values derived previous work. estimated degree contamination ranges between 2 ${{\ \rm per\ cent}}$ 12 cent}}$. results show that ml-moc is reliable segregate stars.