Comparison of K-Nearest Neighbor and Artificial Neural Network Methods for Human Development Index Classification in Sumatra Island

Abstract

The Human Development Index (HDI) serves as a fundamental metric for evaluating quality of life across regions, encompassing health, education, and living standards. This study investigates the comparative performance of K-Nearest Neighbor (KNN) and Artificial Neural Network (ANN) algorithms for classifying HDI categories across districts and cities in Sumatra Island. Utilizing secondary data from 154 regencies/cities obtained from the Central Statistics Agency, this research employs comprehensive preprocessing techniques including data cleaning, transformation, and normalization. The methodology implements three distinct data partitioning schemes (60%-40%, 70%-30%, and 80%-20%) with KNN evaluated at K values of 3, 5, and 7, while ANN utilizes multi-layer feedforward architecture.