中文摘要：

针对多种数据包络分析（Data Envelopment Analysis,简称DEA）模型会产生不同绩效评价结果的问题,提出基于Gini准则科学地融合各DEA模型结果的方法。首先基于Gini准则定义信息纯度以衡量各DEA模型结果的确定性并赋予权重,然后通过加权融合最终得出客观唯一的综合效率。此外,根据评价者的偏好信息或先验知识,进一步提出交互式多DEA模型-Gini准则方法。以前学者仅从单一角度出发选择DEA模型评价高校的运营绩效,考虑到从不同角度出发的多种DEA模型可以给出高校更加全面客观的运营绩效评价,利用以上方法对2011年国内25所理工类高校的运营绩效进行了实证分析,结果验证了以上方法可以合理有效的衡量各高校的运营绩效表现,对于高校运营绩效的评价研究具有实际指导意义。

英文摘要：

To deal with the problem that multiple DEA models might generate different performance evaluation results,an approach based on the Gini criterion is proposed by fusing the results together scientifically.Firstly,the purity of information based on the Gini criterion is defined to measure the certainty of the results,and then the weights are given to each DEA model.A unique set of comprehensive efficiencies can be obtained by weighting the results of each DEA model and the corresponding model weight.Moreover, considering the preference and priori information of estimators,interactive multiple DEA models and Gini criterion approach are further proposed.Measuring the performance of higher education colleges is one of important applications in DEA,and previous researches study the problem just choose the evaluation DEA model from a single point of view.It would be more comprehensive and objective if multiple DEA models that are from multiple points of view are used.Consequently,applying the proposed method to evaluate the performance of 25science and engineering universities in 2011,and the empirical results show that the approach is reasonable and effective,which has practical significance in the research of measuring the performance of higher education colleges.