针对DEA—DA模型在判别过程中对权重无任何限制,导致分析结果与实际状况产生偏差的问题,尝试构建主成分约束锥,将DEA—DA模型与主成分模型各自优势结合,提出了带有主成分约束锥的DEA—DA模型,为以“高维数据组,且变量之间存在相关关系”为研究对象的DMUs分类判别提供更加科学、客观、有效的方法。
Since there are no limitations to the weight in the identification process of DEA-DA model, the results of the analysis are often different from the actual situation of the problem. This paper is intended to build the main components of cone bound to combine the respective strengths of DEA-DA model and principal component model, and proposes the DEA-DA model with the main components of cone-bound to make the method for discriminate classification of DMUs which set "high-dimensional data sets, and the relationship between variables" as the object of the study more scientifically, objectively and effectively.