项目名称: 不确定信息环境下的网络DEA模型分析方法及应用研究
项目编号: No.71301080
项目类型: 青年科学基金项目
立项/批准年度: 2014
项目学科: 管理科学
项目作者: 朱卫未
作者单位: 南京邮电大学
项目金额: 23万元
中文摘要: 数据包络分析(DEA)作为效率评价领域的一种重要方法被广泛应用于各个领域,目前相关研究已扩展到揭开"黑盒"的网络DEA。网络DEA在实践应用中难以处理不确定性信息问题。本课题从区间数据在两阶段决策单元效率评价问题入手,以区间效率来表征评价结果。对强弱序数型、乘子权重置信限制型、模糊型的不确定性数据,通过转换为区间数据或带参数的区间数据进行处理;对随机数据,根据先验信息或概率分布函数通过证据推理理论转化为模糊数据。将不确定信息问题统一到区间数据的处理框架。对数据缺失或无效,使用类似方法解决变量传递过程中的贡献度问题,进行效率评价和数据的区间补完。经结构异化和多阶段扩展,最终将不确定信息问题从两阶段推广至网络DEA。有关成果应用于我国通信运营服务企业的效率评价实践,为企业各级管理者应对政策和竞争环境的变化提供有价值的决策信息,一定程度上也将推动网络DEA方法从理论层次走向更为广阔的应用层面。
中文关键词: 网络结构;不确定数据;DEA;模型拓展;应用
英文摘要: As an important method of performance evaluation, Data Envelopment Analysis(DEA) is widely used in many fields. Nowadays,the research of DEA is extended to network structure DEA, which opens the black box of efficiency analysis. In practise, Network DEA can not deal with most problems of uncertain information. This project researches into two-stage network DEA with interval data and presents a method to evaluate the efficiency of Decision Making Unit (DMU) with interval efficiency.To deal with uncertainty such as strong or weak ordinal data, assurance region restrictions,fuzzy data,we transform them into interval datas.To deal with random data, we transform them into fuzzy data by Dempster-Shafer evidence reasoning Theory(DST) with the presentation of prior information or probability distribution function. So we can solve the problem of uncertain information with the framework of interval data. We also use a similar method to solve the contribution degree of variables during production to performance evalution and data complement dealing with invalid data or null data.By structure alienating and multi-stage extension, we extend the two-stage network DEA model into network DEA with uncertain data.The achievement of this research will be applied to evaluate the performance of telecommunication service operators of
英文关键词: Network structures;Uncertainty;DEA;Extension research;Applications