Study on spatial clustering of urban function partition

As an important means of spatial data mining, spatial clustering has been applied to many fields at present. This paper presents a new method of spatial clustering based on Huffman tree. Using this method, we make quantitative analysis of urban function partition. Moreover, the method has been implemented and applied in a case study in Wuhan city. Simulation experiments pointed out that the method enables people to reason effectively about the law of economical macro distribution, which helps them to mine the hidden available knowledge from mass spatial data of economy that best satisfy their desires.