<![CDATA[Customer segmentation is an important step in supporting marketing strategies on E-Commerce platforms. This study aims to cluster customers based on their characteristics and transaction behavior using the CLARANS (Clustering Large Applications based upon Randomized Search) algorithm. The dataset used consists of E-Commerce customer attributes, including age, average transaction value, total orders, customer loyalty, and churn risk. The research stages include data collection, data cleaning, feature engineering, exploratory data analysis (EDA), algorithm implementation, and clustering evaluation. The evaluation was conducted using Silhouette Score, Davies–Bouldin Index, and Calinski–Harabasz Index, and benchmarked against K-Means and Hierarchical Clustering methods. The results show that the CLARANS algorithm provides the best performance with a Silhouette Score of 0.381991, Davies–Bouldin Index of 1.061123, and Calinski–Harabasz Index of 3458.564. These findings indicate that CLARANS is capable of producing more compact and well-separated clusters, making it effective for customer segmentation in E-Commerce data]]>
Copyrights © 2026
