Comparison of Density-Based Spatial Clustering of Applications with Noise (DBSCAN), K-Means and X-Means Algorithms on Shopping Trends Data

Vina Wulandari (Universitas Islam Negeri Sultan Syarif Kasim Riau, Indonesia)
Yulia Syarif (Universitas Islam Negeri Sultan Syarif Kasim Riau, Indonesia)
Zhevin Alfian (Universitas Islam Negeri Sultan Syarif Kasim Riau, Indonesia)
Muhammad Adil Althof (Sivas Cumhuriyet University, Turkey)
Maylina Mufidah (Sakarya University, Turkey)

Indonesian Journal of Applied Technology and Innovation Science
Vol. 1 No. 1 (2024): IJATIS February 2024

This study extensively compares the efficacy of three clustering algorithms of DBSCAN, K-Means, and X-Means in analyzing shopping trend data, utilizing the Davies-Bouldin Index (DBI) for group validity assessment. The dataset, sourced from Kaggle.com, encompasses various customer attributes. Results indicate that the DBSCAN algorithm demonstrates superior cluster validity, outperforming K-Means and X-Means. Specifically, with an Eps value of 0.3 and MinPts value of 3, DBSCAN achieves an optimal DBI value of 0.1973. K-Means follows with a DBI value of 2.2958, and X-Means attains its best value (2.5663) with k=3. This research underscores the pivotal role of clustering algorithms in understanding shopping trends and customer preferences, offering valuable insights into their comparative performance.