<![CDATA[This study aims to cluster traffic violation data recorded by CCTV in the Sidoarjo area using the K-Means Clustering algorithm. The dataset used in this study was obtained from Sidoarjo Police, covering 43,055 traffic violation records in the period January 2023 to July 2024. The CRISP-DM approach is applied to ensure a systematic research flow, starting from problem understanding, data collection, to result evaluation. After the data selection and transformation stage, the dataset was processed into 14,386 data. Clustering was performed to divide violations into three categories based on severity, namely high, medium, and low. Evaluation of cluster quality using Silhouette Score showed the best result with a value of 0.9916 at k=9, indicating optimal cluster formation. The clustering results showed that the highest violation occurred in the category of “not using a seat belt†with 8,710 cases, while the moderate violation involved “not wearing a helmet†with 5,522 cases. This study confirms the effectiveness of the K-Means algorithm in clustering traffic violation data and provides valuable insights for the Sidoarjo Police Traffic Unit in designing more efficient traffic violation reduction programs.]]>
Copyrights © 2025
