基于拓展聚类模型的区域创新能力层级划分研究  

Research into Classification of China's Regional Innovation Capability Based on Classification of Extended Clustering Model

作　　者：王德青[1,2] 朱建平[1,2] 

机构地区：[1]厦门大学经济学院统计系,福建厦门361005 [2]厦门大学数据挖掘研究中心,福建厦门361005

出　　处：《经济经纬》2014年第1期8-13,共6页Economic Survey

基　　金：国家社会科学基金项目(11BTJ001);国家自然科学基金青年项目(710201139);全国统计科学研究计划重大项目(2012LD001)

摘　　要：为了对中国创新能力科学划分以深层次挖掘区域之间创新能力的有机联系,笔者在剖析现有层级划分方法优点与不足的基础上进行模型拓展,应用拓展的聚类模型对我国的区域创新能力进行静态与动态分类,并对分类结果的显著性进行检验。对比分析不同地区创新发展模式发现,创新能力东强西弱的阶梯分布格局仍存在,必须分集团而不是整齐划一的制定和实施创新政策。In order to classify China＇ s innovation capability and mine the deep organic connection among regions＇ innovation capabili- ty, the authors extend models based on the advantages and defects of the summation of existed classification methods, and apply im- proved clustering models to classify China＇ s regional innovation capability both dynamically and statically. Making significance tests and comparing innovation development mode of different regions, the authors find that there still exists the innovation step distribution advanced in the east and weak in the west. Accordingly, innovation policy should be made and implemented separately rather than uniformly.

关 键 词：区域创新能力 聚类分析 主成分分析 面板数据 层级划分 

分 类 号：F061.5[经济管理—政治经济学] C81[社会学—统计学]