Aradi Sebayang
Universitas Pembangunan Panca Budi

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

ANALISIS KOMPARASI K-MEANS DAN K-MEDOIDS DALAM PEMETAAN WILAYAH PRIORITAS DISTRIBUSI BBM BERSUBSIDI SUMATERA UTARA Ade Iskandar; Aradi Sebayang; Tengku Didi Ferdillah; Toni Prabowo; Muhammad Fuad Hafiz; Muhammad Zainal Arifin Pohan; Muhammad Syahputra Novelan
JOURNAL OF SCIENCE AND SOCIAL RESEARCH Vol. 9 No. 3 (2026): June 2026
Publisher : Smart Education

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54314/jssr.v9i3.6220

Abstract

Abstract: The inequality in the distribution of subsidized fuel (BBM) is a strategic issue in North Sumatra Province, influenced by the volume of motorcycles, cars, buses, and trucks across 33 Regencies/Cities. This research aims to map priority distribution areas using clustering techniques by comparing the performance of the K-Means and K-Medoids algorithms. Real data on the number of vehicles from 2025, sourced from BPS North Sumatra Province, serves as the primary variable. Evaluation results using the Silhouette Score indicate that the K-Means algorithm demonstrates superior performance with a score of 0.63, compared to K-Medoids which only reached 0.09. K-Means successfully identified Medan City as an extreme outlier requiring independent distribution policies, whereas K-Medoids experienced overlap among smaller regional clusters. These findings provide empirical recommendations for policymakers to ensure that subsidized fuel quota allocations are more accurately targeted. Keywords: K-Means; K-Medoids; Subsidized Fuel; Clustering; North Sumatra.   Abstrak: Ketimpangan distribusi Bahan Bakar Minyak (BBM) bersubsidi merupakan isu strategis di Provinsi Sumatera Utara yang dipengaruhi oleh volume kendaraan motor, mobil, bus, dan truk di 33 Kabupaten/Kota. Penelitian ini bertujuan memetakan wilayah prioritas distribusi menggunakan teknik clustering dengan membandingkan performa algoritma K-Means dan K-Medoids. Data riil jumlah kendaraan tahun 2025 dari BPS Provinsi Sumatera Utara digunakan sebagai variabel utama. Hasil evaluasi menggunakan Silhouette Score menunjukkan bahwa algoritma K-Means memiliki performa lebih unggul dengan skor 0,63 dibandingkan K-Medoids yang hanya mencapai 0,09. K-Means berhasil mengidentifikasi Kota Medan sebagai extreme outlier yang memerlukan kebijakan distribusi mandiri, sementara K-Medoids mengalami tumpang tindih (overlap) pada klaster daerah kecil. Temuan ini memberikan rekomendasi empiris bagi pengambil kebijakan agar alokasi kuota BBM subsidi lebih tepat sasaran. Kata kunci: K-Means; K-Medoids; BBM Bersubsidi; Clustering; Sumatera Utara.