Classification of the Mainstay Economic Region Using Decision Tree Method

Abstract

The development of the region cannot be separated from the concept of economic growth and the determination of the mainstay region as a regional center that is expected to have a positive impact on economic growth to the surrounding regions. In fact, the determination of the mainstay region is a difficult thing to do. Some cases of the determination of the mainstay region are mostly on the basis of the prerogative rights of the policy makers without carefully seeing the achievements of the development of a region. The objective of this study is to develop a classification model of the mainstay economic region using computational techniques. The decision tree methods of NBTree and J48 are used in this study and combined with Klassen typology. The results of this study show that J48 algorithm has better accuracy than NBTree in the formation process of decision tree. The accuracy of J48 is higher than NBTree i.e. 68.96%. The comparative result of the classification of the mainstay economic region between Klassen and J48 shows that there is a shift in the class position of the development quadrant. In Klassen classification, there are three regions that are categorized into the mainstay regions with advanced development and rapid growth (K1). Meanwhile, J48 results show that there is no region categorized into K1. However, the mainstay economic region on J48 is based on the level of development with the level below K1, i.e. K2. J48 classification results show that there are ten regencies that are categorized into the mainstay economic regions, namely Biak, Regency of Jayapura, Jayawijaya, Kerom, Merauke, Mimika, Nabire, Ndunga, Yapen, and the Municipality of Jayapura.