KEANGGOTAAN GUGUS TERBUKA GANDA DENGAN METODE PEMBELAJARAN MESIN HDBSCAN

• Main Author: Chikmah
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/69335
• Institution: Institut Teknologi Bandung

A binary cluster system is a system between two cluster that are close to each other and interact dynamically. Studies of binary clusters have been carried out for binary clusters in the Large Magellanic Cloud, the Small Magellanic Cloud, and the Milky Way. Research on binary clusters can provide crucial information about the mechanisms of cluster formation and evolution. In this Final Project, the determination of cluster membership in pairs of binary open clusters is carried out using the Hierarchical Density-Based Spatial Clustering of Applications with Noise (HDBSCAN). HDBSCAN is a machine learning algorithm with a hierarchical approach to automatic grouping based on cluster stability. The membership of stars in a cluster is then determined by the magnitude of the probability of membership. The data used in this Final Project is Gaia DR3 data with a cluster magnitude less than G ~ 18 mag. The pairs reviewed in this Final Project are three binary clusters with significant separation differences, Berkeley 4-King 16, NGC 5617-Trumpler 22, Basel 18-Collinder 271. HDBSCAN results showed a pretty good clustering for the Basel Berkeley 4-King 16 and 18-Collinder 271 binary clusters. In the NGC 5617-Trumpler 22 cluster there is a cluster projected in the same field as NGC 5617, namely Pimis 19. HDBSCAN has not been able to separate the members of Pismis 19 from NGC 5617. The results of this membership were then compared with the catalog of Dias etal. (2021) and catalog of Cantat-Gaudin & Anders (2020). It was also reviewed the interaction between cluster pairs through a density map. In this Final Project, it was found that there were cluster candidates that had not been listed in the reference catalog and needed to be reviewed further.