Development Potential and Future Direction of National New Areas——A GRNN Approach
用广义回归神经网络评估了17个新区的18项指标，发现上海、深圳、广州、天津的新区潜力最大，沿海新区比内陆更开放，但重庆例外。
In this study, GRNN(general regression neural network) approach is applied to evaluate the development potential of 17 new areas(the 9 established national new areas and 8 new areas with strong possibility to be granted as national importance). The evaluation system consists of 18 indicators of 4 broad aspects, reflecting economic gross scale,economic openness, innovation capability and transportation communication networks. The result shows that:(1)Significant disparities of development potential exist among the new areas, among them those in Shanghai, Shenzhen,Guangzhou and Tianjin gain overwhelming advantages; generally, coastal new areas are more open than inland new areas; while Chongqing, as an exception, not only has high level of openness, but has great advantage in gross scale,which makes its new area comparable to those in Shanghai, Shenzhen, Guangzhou and Tianjin.(2) Cluster analysis based on the GRNN result suggests that, the future development paths of the 17 new areas can be classified into: internationally competitive area, national center, regional center and special strategy oriented area. At the end the article, based on their strategic goals and GRNN development potential evaluation results, differential function of these areas is discussed, and further suggestions on their future development approached is given.