摘要
针对于主成分分析方法在区域竞争力研究中所表现出的局限性与不足,本文采用了一种基于相关系数标准差(CCSD)的多属性评价方法。该方法综合考虑了指标属性值的相关程度和偏差大小对评价结果的影响,并通过构建CCSD模型将其影响反映到指标权重的确定上,从而能够有效地克服主成分分析方法和传统的多属性评价方法在评价实施过程中属性权重确定的主观性。CCSD多属性评价方法在指标权重确定中消除了指标间的相关性影响,因此在指标选择上不再受统计量的局限,从而能够通过指标数据确定指标权重,依据客观数据得到评价结果。本文应用该方法对我国31个省市自治区的经济竞争力进行了对比研究,发现其评价结果与实际经济发展现状相一致,从而证实了该方法的有效性和实用性。
Addressing the defect and limitations of principal component analysis method used for regional competitiveness research,this paper uses a correlation coefficient standard deviation model for multi-attribute evaluation method.The method takes into account the impacts of the correlation and deviation of corresponding index attribute values on the evaluation results,and reflects them in the determination of weights.The CCSD method eliminates the correlation of index to determine the index weights,so the index selection no longer subjects to the limitations of statistics.The index data are used to get the objective weights and evaluation results.On evaluating the economic competitiveness of 31 provinces in China,the evaluation results show consistency with the status quo of the actual economic development,which indicate the method is effective and practical.
出处
《运筹与管理》
CSSCI
CSCD
北大核心
2012年第2期249-255,共7页
Operations Research and Management Science
基金
国家自然科学基金资助项目(70801022)
关键词
运筹学
多属性评价
CCSD模型
区域竞争力
主成分分析
operations research
multiple attribute evaluation
CCSD model
regional competitiveness
principal component analysis
