Hierarchical Clustering in Astronomy

TL;DR

This paper reviews the development and application of hierarchical clustering in astronomy, highlighting its role in revealing celestial structures and automating object classification across various scales.

Contribution

It provides a systematic review of hierarchical clustering methods in astronomy, clarifying their conditions, limitations, and applications in different astronomical research areas.

Findings

Hierarchical clustering reveals intrinsic structures of celestial systems.

It automates classification of large astronomical datasets.

Clarifies conditions and limitations of clustering methods.

Abstract

Hierarchical clustering is a common algorithm in data analysis. It is unique among many clustering algorithms in that it draws dendrograms based on the distance of data under a certain metric, and group them. It is widely used in all areas of astronomical research, covering various scales from asteroids and molecular clouds, to galaxies and galaxy cluster. This paper systematically reviews the history and current status of the development of hierarchical clustering methods in various branches of astronomy. These applications can be grouped into two broad categories, one revealing the intrinsic hierarchical structure of celestial systems and the other classifying large samples of celestial objects automatically. By reviewing these applications, we can clarify the conditions and limitations of the hierarchical clustering algorithm, and make more reasonable and reliable astronomical discoveries.