<![CDATA[This study aims to assess the performance of single linkage, complete linkage, and average linkage hierarchical clustering algorithms in grouping cities used as inflation benchmarks in Indonesia into clusters based on sectoral inflation patterns. The data utilized are 150 regencies/cities divided into 11 sectors that drive inflation, identified by BPS Indonesia. Prior to clustering, a distance analysis using Euclidean distances was conducted to measure similarity between regions. Evaluation of the optimal number of clusters was conducted by applying the stability measure approach (APN, AD, ADM, and FOM), which showed that creating five clusters produced the most stable results. The results of the analysis revealed that the single linkage approach had the lowest within-cluster to between-cluster standard deviation ratio compared to the other two approaches, which revealed a greater level of homogeneity between the clusters. From an economic perspective, this clustering pattern revealed impressive differences in sectoral inflation pressures between provinces, even between cities within a province. Consequently, the single linkage method is proposed as the optimal method for identifying spatial variations in sectoral inflation in Indonesia.]]>
Copyrights © 2026
