Implementasi machine learning dalam pengelompokkan provinsi di indonesia berdasarkan data pencemaran lingkungan hidup

Abstract

Environmental pollution is a crucial issue that needs serious attention. The increasing world population will also increase the level of environmental pollution (Mittal & Mittal, 2013) especially in developing countries (Remilekun Adeuti, 2020) This is also the case in Indonesia. Therefore, this research aims to find out which provinces in Indonesia have a high level of pollution by clustering provinces based on environmental pollution data. The methods used in this research are K-Medoids, K-Means, and Fuzzy C- Means as well as Complete Linkage and Ward's Linkage for Agglomerative Hierarchy. The results show that the K-Medoids method is the best method produces 3 clusters, namely clusters with high average pollution of 11 provinces, clusters with average pollution of 12 provinces, and clusters with low average pollution of 11 provinces.