Discovery of rules in urban public facility distribution based on DBSCAN clustering algorithm
Xinyan Li, Deren Li · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2007
Recently Spatial Data Mining (SDM) has been recognized as a powerful technology that can complement traditional GIS to facilitate urban planning and management since it can be used to discover interesting, implicit knowledge from spatial database. DBSCAN spatial clustering algorithm as a SDM method is able to effectively discover clusters of arbitrary shape in large database with noise points. In this paper we applied this algorithm to detect distribution patterns of urban public facilities in a developed city, including primary school, high school and commercial facilities. Both qualitative and quantitative analysis were carried out to investigate how to determine optimal values of input parameters for DBSCAN algorithm, and the distribution patterns of public facilities were assessed against urban planning design standard using the algorithm.