改进的DEA模型及其在上市物流企业中的应用

数据包络分析(DEA)是对具有多投入、多产出决策单元(DMU)进行效率评价的一种非参数法,其研究领域一类重要问题是探讨DEA改进科学性。针对J-K模型提供的优化方案中常会出现以巨大的投入换得产出微弱增加的问题,将容忍度概念引入DEA法,提出一种改进的J-K数据包络分析模型,并对修改前后的两个模型进行比较研究,结果表明:改进的J-K数据包络分析模型既克服了传统DEA模型对投影要求的绝对化(必须是所有指标全部改进),也排除了J-K模型提供不经济方案的可能性(以巨大的投入换得产出微弱增加)。因此,应用改进的J-K数据包络分析模型可得到更加合理的分析结果。基于典型上市物流企业数据也验证了改进的J-K数据包络分析模型的合理性。

A Modified Model of DEA and Its Application in Listed Companies of Logistics

Data envelopment analysis(DEA) is a non-parameter method particularly suitable for dealing with the efficiency evaluation of DMU of a multi-input and multi-output nature.The issue of primary importance in DEA study is the embetterment of its scientificity.The optimization results,which the J-K model provides,often cause huge cost of input in exchange for a slight increment of output.For this reason,this paper introduces the concept of tolerance and proposes an amended J-K data envelopment analysis model.Furthermore,the comparative study between J-K model and the modified one is discussed.Our study shows that the modified model of DEA not only can overcome the absolute projection of traditional DEA model(All the indicators must be improved),but also exclude the possibility of non-economic program which the J-K model provides(huge cost of input in exchange for a slight increment of output).Therefore,more reasonable results can be obtained by using the improved J-K model.Finally,based on listed company of logistics,the model provided in the paper is justified by empirical analysis.