论文部分内容阅读
数据包络分析(DEA)是一种评价具有多投入、多产出决策单元的相对效率的线性规划方法.在现实世界中,决策单元有时呈现出由多个独立子系统构成的复杂并联网络系统,各子系统的投入/产出之和构成了系统的总投入/产出.目前,用于评价这种具有并联网络生产系统相对效率的模型主要有三种:网络DEA模型、多部门DEA模型和关联DEA模型.现有这些模型的基本特性和相互关系存在着不足,即子系统的效率分解和优化指数不唯一.为解决这一问题,提出了改进的并联DEA模型,并采用加拿大银行系统实例来说明所提出模型的合理性和有效性.
Data envelopment analysis (DEA) is a linear programming method to evaluate the relative efficiency of multi-input and multi-output decision-making units.In the real world, decision-making unit sometimes presents complex parallel network system composed of many independent subsystems , The sum of input / output of each subsystem constitutes the total input / output of the system.At present, there are mainly three models for evaluating the relative efficiency of this parallel networked production system: network DEA model, multisectoral DEA model and Related DEA model.The existing characteristics of the existing models and their interrelationships are not enough, that is, the efficiency decomposition and optimization index of the subsystems are not unique.In order to solve this problem, an improved parallel DEA model is proposed and the Canadian bank system instance To illustrate the proposed model is reasonable and effective.