Aplikasi Prediksi Kepadatan Penduduk Menggunakan Decision Tree

Abstract

The continuous increase in population growth requires an analytical system capable of providing predictive information to support regional development planning. This study aims to develop a population density prediction application using the Decision Tree algorithm by utilizing official population data from the Central Statistics Agency (BPS) covering population and area. The research process is carried out through several stages, namely problem identification, data collection, system design, model implementation, and application evaluation. The Decision Tree algorithm was chosen because it is able to provide a decision tree structure that is easy to interpret and effective for tabular data-based classification. Attribute separation measurements were carried out using the Gini Index, Entropy, and Information Gain to determine the best attribute to form a node. The analysis results show that the population attribute is the most influential variable in determining the density category, with the highest Information Gain of 1.5269. The model produces clear classification rules, such as low density categories for areas with a population of less than 3 million, medium for 3–10 million, and high for more than 10 million people. The application evaluation shows that the system is able to run stably and provides accurate prediction results according to the data pattern. The application developed is expected to assist local governments in monitoring and anticipating changes in population density as a basis for data-driven planning.