Clustering Musik Rock Menggunakan Algoritma K-Means dan K- Medoids

Clustering is a data analysis technique used to group objects with similar attributes or
characteristics. The goal of clustering is to uncover hidden structures or patterns in data
without prior label or classification information. This study implements K-Means and KMedoids algorithms to analyze the "Top Hits Spotify from 2000-2019" dataset, clustering
rock music based on audio track attributes to produce clusters and compare the results with
high, medium, and low comparisons for each audio track feature. The analysis reveals
significant variations in music characteristics and genres across different clusters. The KMeans algorithm generates two clusters: Cluster 1 (142 members) is dominated by rock,
pop, and dance/electronic genres with high popularity and danceability, while Cluster 2 (83
members) includes rock, pop, and metal genres with prominent energy attributes. The KMedoids algorithm produces five clusters with higher diversification, where Cluster 5 (77
members) is dominated by rock, pop, and metal genres with high loudness, and Cluster 1
features rock, pop, hip-hop, and metal genres with the highest mode values, indicating
consistent use of major or minor tones. These results reveal differences in genre preferences
and music attributes, reflecting the complexity and diversity of the analyzed music data.