Cluster analysis on city real estate market of China: based on a new integrated method for time series clustering

AbstractAfter the reform of urban housing system in 1998, China real estate market had a rapid growth in recent years, while house price was increasing sharply. Using the House Price Indices of 70 cities in China from CREIS (China Real Estate Index System), we found that the house price of each city had an upward tendency with some certain stages. However, different cities also had their distinctive features. In this paper, a new integrated method for time series clustering is employed to do cluster analysis on city real estate market of China. The time series are firstly divided into several stages mainly based on the changes in government policy using wavelet analysis with expert experience. Then the variables that describe the character of each stage such as average growth rate and volatility are used as attributes of each city. Consequently, DBScan algorism for normal clustering can be used and the results show that there are several categories of growth modes of city real estate markets while the macro-control policies had different effect on each category