Cluster analysis: an overview of traditional and novel approaches for health researchers

Cluster analysis has gained increased importance with the rapid expansion of artificial intelligence (AI) and data-driven research. As a central method in unsupervised machine learning, it can reveal hidden patterns and identify homogenous subgroups within data without requiring prior grouping labels. This approach not only deepens clinical understanding but also offers an effective means of supporting cluster-informed， tailored care to improve patients’ disease outcomes. The purpose of this paper is to provide an overview of cluster analysis, covering its historical and conceptual foundation, commonly used cluster analysis methods, the role of dimensionality reduction (DR) techniques, and practical steps for implementation.