Agus Maman Abadi, Muhammad Fauzan, Caturiyati Caturiyati, Bintang Wira Mahardika, Muhammad Farhan Yazid, Yoswan Septianto
The Yogyakarta Special Region faces a decline in water quality, with river conditions as a significant factor. Previous studies highlight clustering techniques use K-Means and Fuzzy C-Means (FCM) as effective for water quality assessment, but comparison between these algorithms has not been conducted. This study investigates: How do K-Means and FCM algorithms perform and which algorithm provides optimal outcomes in clustering 10 river water quality in the Yogyakarta Special Region from 2020-2023 based on pollution index? The hypothesis suggests K-Means outperforms FCM due to its centroid-based optimization. Evaluation using the Calinski-Harabasz Index (CHI), Davies Bouldin Index (DBI), and Dunn Index (DI) showed that K-Means has higher CHI and DI values and lower DBI values, indicating better-defined clusters than FCM. The clustering results identified Cluster 2 as the group with the highest pollution level, followed by Clusters 4, 1, 0, and 3. Cluster 2 has the highest pollution index value consistently, indicating the urgency of priority intervention. These findings can be used to formulate river pollution mitigation policies. With these results, the study not only demonstrates the optimal algorithm in water quality clustering, but also provides data-based recommendations for more effective environmental management in Yogyakarta and other areas with similar characteristics. © (2025), (UIKTEN - Association for Information Communication Technology Education and Science). All rights reserved.
Department of Mathematics Education, FMIPA, Yogyakarta State University, Karangmalang Campus, Colombo Street No.1, Yogyakarta, 55281, Indonesia
