본문 바로가기
  • Home

Simulating Land Use Change Using Decision Tree and SVM Model : A Case Study of North Korea’s City after the Unification

Seok Hwan Won 1 Hwang Chul Sue 1

1경희대학교

Accredited

ABSTRACT

This study aims to estimate land-use changes in North Korea’s city after the unification through simulation. The simulation analysis consists of exploration of urban patterns, validation of simulated results, and analysis of the land-use changes in the future. The estimation accuracy of land use change was measured by polynomial regression analysis, decision tree analysis and support vector machine analysis. On this basis, suitable methodologies were selected for the modeling of the urban pattern search. The hybrid model uses the ‘C5.0’, a kind of decision tree algorithm and the ‘RBF kernel’, a kind of support vector machine algorithm. The land use change of North Korea’s city (Nampo City) was estimated based on this model. The scenario of newly constructing roads and railroad stations, and developing new urban areas was used in the simulation. The simulated results showed that the urban sprawl impact through improvement of traffic accessibility and the development of new urban areas could be linked to the region and other regions. This study could provide a strategic guidance for policy when the unification of Korea will be on the process.

Citation status

* References for papers published after 2023 are currently being built.