Exploring Clustering Algorithms for Parkinson's Disease Data: A Comparative Analysis

Abstract

Clustering, an essential analytical approach utilized in data mining, encompasses the act of grouping alike data items into clusters. It is crucial to note that the clustering outcome is significantly impacted by the employed clustering algorithm. This research paper presents a thorough analysis of various clustering algorithms, such as k-means, hierarchical, and DB-scan clustering algorithms, among others, while simultaneously scrutinizing their strengths and limitations. Within each algorithm type, the computation of the distance between data objects and cluster cantres is executed in every iteration, which inevitably poses a challenge to the efficiency of clustering. This paper provides an extensive summary of the fundamental techniques and highlights the associated challenges with clustering algorithms, such as recall, precision, and f-measure, to produce superior outcomes under diverse circumstances. The paper concludes with a discussion of the results obtained from a high-dimensional dataset of Parkinson's disease