<![CDATA[Public transportation is a primary mode of intercity travel in Aceh Tamiang, yet operators face challenges in balancing fleet availability with fluctuating ticket demand. This study aims to analyze ticket booking patterns for buses and minibuses (Hiace) using the Partitioning Around Medoids (PAM) clustering method to provide a basis for operational optimization. The research applies the Knowledge Discovery in Databases (KDD) framework to historical booking data from the local Transportation Agency. The PAM algorithm effectively grouped the 15 main routes into three distinct clusters: Cluster 1 (12 routes) with stable-medium demand, Cluster 2 (2 strategic city routes) with high demand, and Cluster 3 (1 inter-province route) with very high demand. These results successfully identify hidden demand patterns, with the final clustering achieving a low Total Cost of 2.0. The findings directly translate into actionable recommendations for differential scheduling and fleet allocation, demonstrating the practical value of data mining in public transport management.]]>
Copyrights © 2026
