摘要
在设计区域创新能力评价初始指标体系的基础上,采用基于主基底分析的变量筛选方法对初始指标体系进行变量筛选,构建了区域创新能力评价体系的简约变量集合,在保证原始变量信息损失尽可能小的前提下,排除所有的冗余信息,避免了变量之间的多重相关性;再利用主成分分析对简约变量集合进行二次降维,构建综合评价模型,对我国30个地区的创新能力进行排序及分析评价,得到了较好的评价结果。
Based on the design of the original evaluation index system of regional innovation capability,the paper selects variables by a variable selection method based on principal basis analysis,to condense the variable set for evaluation of regional innovation capability.The adopted method selects information effectively from the large-scale variable set,while excluding all the redundant variables and reduplicate information on the promise that the loss of original information is minimized.Then principal component analysis is applied to accomplish the second dimension reduction on the condensed variable set.In this way,a comprehensive evaluation model is constructed to rank and evaluate the regional innovation capability of thirty areas in China.
出处
《系统工程》
CSSCI
CSCD
北大核心
2011年第7期34-40,共7页
Systems Engineering
基金
国家自然科学基金创新研究群体科学基金资助项目(70821061)
国家自然科学基金资助项目(70771004)
国家自然科学基金重点资助项目(71031001)
关键词
管理工程(系统工程)
区域创新能力评价
主基底变量筛选法
主成分分析
简约变量集合
Management Engineering
Evaluation of Regional Innovation Capability
Variable Selection Method of Principal Basis Analysis
Principal Component Analysis
Condensed Variable Set
