基于DEA系统潜力损失的决策单元全排序方法  

A Fully Ranking Method for DMUS Based on DEA System Potential Loss

作　　者：王军[1] 刘晶[1] 孙利辉[1] 

机构地区：[1]青岛大学管理科学与工程系,山东青岛266071

出　　处：《河南科技大学学报（自然科学版）》2010年第2期100-104,共5页Journal of Henan University of Science And Technology:Natural Science

基　　金：国家自然科学基金项目(70571040);山东省教育厅科技计划项目(J07WJ09)

摘　　要：DEA模型作为一种非参数评价方法,在很多领域得到广泛应用。传统的DEA方法经常会遇到多个决策单元同为DEA有效而无法区分相对优劣的情况。为了对决策单元进行全排序,本文从系统和相对效率的角度定义了相对潜力和潜力损失的概念,揭示潜力损失和强弱有效性之间的内在关联,并提出一个基于DEA系统潜力损失的决策单元全排序方法,来判定所有决策单元的优劣,最后用一个实例来说明本方法的计算步骤。The data envelopment analysis (DEA) is a nonparametric evaluation method,and has been applied in many fields.When the efficient DMUs exist in high percentage,DEA can not discriminate them completely in most cases.This paper defines DEA system,system potential and potential loss in terms of relevant efficiency,and reveals the essential relationships between the potential loss and the strong-week efficiency.A new fully ranking method for DMUs is proposed based on the potential loss.An example is used to demonstrate the iterative and practicability of the new method.

关 键 词：数据包络分析 决策单元 绩效评价 全排序 

分 类 号：N945.16[自然科学总论—系统科学]