Defesa de dissertação de mestrado
Student: Giulya Souza dos Santos
Program: Astronomy
Title: "Searching for Merging Galaxy Clusters: Exploring Clustering Algorithms in Photometric Surveys to Identify Substructures in Clusters"
Advisor: Prof. Dr. Eduardo Serra Cypriano
Judging Committee
- Prof. Dr. Eduardo Serra Cypriano - Presidente e Orientador - IAG/USP
- Profa. Dra. Claudia Mendes de Oliveira - IAG/USP
- Prof. Dr. Rubens Eduardo Garcia Machado – UFTPR (por videoconferência)
- Prof. Dr. André Luis Batista Ribeiro – UESC (por videoconferência)
Abstract
The identification of galaxy clusters undergoing mergers in observational data requires the search for signatures that often manifest as substructures in the galaxy distribution. In this work, we evaluate and optimize the AME/HDBSCAN algorithms to identify substructures in galaxy systems with a bimodal distribution. Our goal is to develop a method capable of identifying good candidates for merging clusters in observational catalogs that use photometric data. We conducted a study using numerical simulations to adapt the algorithm for this purpose, and the results indicate that its optimized version achieves a purity of 74% in selecting satisfactory detections, in addition to an accuracy above 90%. We conclude that the developed method is effective in identifying well-defined substructures decoupled from the main cluster, being able to select the most reliable detections while rejecting spurious ones.
Keywords: Galaxy Clusters, Merging Clusters, Clustering Algorithms, HDBSCAN