MULTIPLE OPEN CLUSTER MEMBERSHIP USING DBSCAN MACHINE LEARNING METHOD

• Main Author: Christopher Anandito, Michael
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/86257
• Institution: Institut Teknologi Bandung

A binary cluster system refers to a stellar cluster system composed of two closely spaced star clusters that interact and are gravitationally bound to each other. Each pair of binary clusters has its own unique characteristics in terms of formation and evolutionary processes. The understanding gained from studying binary clusters is relevant as it provides crucial information about the formation and evolution mechanisms of star clusters. This Final Project aims to determine the cluster membership within binary open clusters using the machine learning clustering method Density-Based Spatial Clustering of Applications with Noise (DBSCAN). DBSCAN is an unsupervised machine learning model that, based on density, is capable of finding patterns or clusters within data. The Gaia Data Release 3 dataset is used in this Final Project, with a cluster magnitude parameter of less than 18 magnitudes. The focus of this thesis is on three pairs of binary clusters reviewed from the Angelo et al., 2021 and Subramaniam et al., 1995 catalogs, with differences in their interactions: King 16-Berkeley 4, NGC 5617-Trumpler 22. The membership results obtained from this method are compared with the membership results of the Dias et al., 2021 and Cantat-Gaudin & Anders, 2020 catalogs.