Application of Block-Based K-Medoids and Ward's Method to Classify Provinces in Indonesia Based on Environmental Quality Index

Abstract

Environmental quality is an essential aspect of the world's sustainability. This paper presents a combination of several methods to obtain a profile of provinces in Indonesia according to the environmental quality indices. We apply the block-based k-medoids algorithm, Ward's method, and the Hartigan index to determine the number of clusters. Based on Hartigan index suggested classifying 34 provinces in Indonesia into four groups. A comparison of the mean vector of four groups using multivariate analysis of variance concluded that three areas have the best environment quality index and eight regions require serious attention to improve their environmental quality methods.