Application of the Ant Colony Algorithm in the Identification of Globular Clusters

Hollander, Jördis (2022) Application of the Ant Colony Algorithm in the Identification of Globular Clusters. Bachelor's Thesis, Artificial Intelligence.

Preview	Text final-submission-application-of-the-ant-colony-algorithm-in-the-identification-of-globular-clusters-j-hollander-s2956543.pdf Download (21MB) | Preview
	Text toestemming.pdf Restricted to Registered users only Download (96kB)

Abstract

Globular clusters (GCs) are stellar agglomerates of about 10000 to 100000 stars and are an interesting ground to study stellar evolution. In this paper, a pipeline for the identification of GCs is developed. This pipeline excludes candidate regions through the use of a blob-detection technique. The remaining regions are then processed by the Ant Colony random-walk algorithm to provide information on stellar density in the form of pheromone values. Finally, these results are fed into a gravity-inspired clustering algorithm to determine potential GCs. The results of the full pipeline identifies 41 clusters of which 27 could be identified as known stellar structures. These clusters are (31.7%) 13 GCs, (12.2%) 5 Open Clusters, (9.8%) 4 Galaxies, (4.9%) 2 Dwarf Galaxies, (2.4%) a Molecular Cloud, (2.4%) an Absorption Nebula, and (2.4%) an Emission Nebula. In addition, it finds (34.1%) 14 clusters that do not correspond to a known stellar structure. From these results, the blob-detection operates as an effective exclusion criteria but with the current parameters it does not yet maintain all known GCs. While the pipeline does not identify the majority of the GCs that exist, for those it can identify, it pinpoints their locations accurately. Further research in tuning the parameters and the behavior of the ants is expected to increase the GCs identified by the pipeline. Further refinement of this process could likely make this pipeline robust.

Item Type:	Thesis (Bachelor's Thesis)
Supervisor name:	Jaeger, H. and Balbinot, E.
Degree programme:	Artificial Intelligence
Thesis type:	Bachelor's Thesis
Language:	English
Date Deposited:	24 May 2022 10:21
Last Modified:	24 May 2022 10:21
URI:	https://fse.studenttheses.ub.rug.nl/id/eprint/27086

Actions (login required)

View Item