Spatiotemporal modeling of 2000–2009 financial development in Yangtze River Delta using support vector machine and clustering analysis

Over the past decade, the Yangtze River Delta has witnessed a financial takeoff. Its mechanism need be understood. The paper attempts to integrate a number of analytic methods to model the financial development of 16 cities in the region from 2000 to 2009. A set of indicators describing the economic environment, foreign trade conditions, banking and insurance industry, and urban development are first identified, with the financial interrelations ratio being selected to represent financial development. Then, the time series data of each city are used to build a temporal model under the theoretical framework of SVM. In the model, the method of cross-validation is introduced for choosing the best parameter of the kernel function and the penalty coefficient to be used later in model training and regression. In the final analysis, the time series data of each city are analyzed through clustering, showing the dynamics of the cities during 2000-2009.