ANALISIS POLA PRODUKSI SAMPAH MENGGUNAKAN K-MEANS, DBSCAN, DAN HIERARCHICAL CLUSTERING

Abstract

The increasing volume of waste without data-driven management creates challenges in designing effective environmental policies. This study aims to identify waste production patterns using clustering approaches through three algorithms: K-Means, DBSCAN, and Hierarchical Clustering. The dataset consists of waste volume, population density, and collection frequency variables that were normalized prior to analysis. K-Means was applied for centroid-based segmentation, DBSCAN for density-based clustering and anomaly detection, and Hierarchical Clustering for analyzing structural relationships among regions. The results indicate that all three methods successfully grouped regions based on waste production characteristics using complementary approaches. K-Means produced stable segmentation, DBSCAN detected extreme-pattern regions, and Hierarchical Clustering provided hierarchical cluster structures. This research contributes to data-driven waste pattern mapping to support more effective and sustainable environmental decision-making.   Keywords: Waste Production Patterns, Clustering Analysis, K-Means Algorithm, DBSCAN Algorithm, Hierarchical Clustering