基于多元统计分析的城市竞争力研究

Research on Urban Competitiveness Based on Multivariate Statistical Analysis

应用多元统计分析方法对34个省会城市和计划单列市的竞争力进行了综合评估.首先设立城市竞争力评价指标,由于指标之间存在相关性,采用因子分析法选取公共因子作为评价的综合指标,然后构造评价函数,根据城市竞争力在各公因子上的得分及其综合排名,分析各城市的发展特色和竞争优劣点.基于因子分析法得到的公因子替代原始变量,分别采用系统聚类分析法和K均值聚类分析法给各城市划分类型,比较两种不同聚类法得出的结果,分析聚类结果可靠性.最后基于综合评估结果给出了提高城市竞争力的合理建议.

This paper made a comprehensive study on the urban competitiveness of 34 provincial capitals and municipali- ties.First, the evaluation index of urban competitiveness was set up.Due to the correlation between the indicators, the factor was used as the comprehensive index of the evaluation,and then the evaluation function was constructed.According to the scores of the city＇s competitiveness on the common factor and its comprehensive rank,the development characteristics of the city and the advantages and disadvantages of competition were analyzed.Based on the factor analysis, the common factors were replaced by the common variables.The clustering types and the K-means clustering analysis were used to classify the cities respectively.The results of the two different clustering methods were compared and the reliability of the clustering results was analyze＆Finally, based on the comprehensive evaluation of the results, some recommendations were put forward to improve the competitiveness of the cities.