论文部分内容阅读
传统数据包络分析模型的生产前沿面构造方法较为复杂,为了解决这一问题,提出了一个新的构建DEA生产前沿面的算法,称为旋转法.阐述了旋转算法的含义.分别从二维和高维角度对传统的四类DEA模型的生产前沿面进行构建,并对算法的理论依据进行了证明.通过实际算例证明,相比起传统的方法,比如极点极方向法、Graham扫描法等,旋转法非常简单而且实用价值很广.最后提供了一个关于该算法在实际中的应用——构建决策单元中含有负值的DEA模型,解决了传统DEA模型无法处理决策单元中含有负值的不足.
In order to solve this problem, a new algorithm for constructing the frontier of DEA production is proposed, which is called the rotation method. The meaning of the rotation algorithm is expounded. From the two-dimensional And the high-dimensional perspective of the traditional four types of DEA model production frontier surface to build, and the theoretical basis of the algorithm is proved.An actual example shows that compared to the traditional methods, such as pole polar direction method, Graham scanning method And so on, the rotation method is very simple and has a wide range of practical value.Finally, it provides a practical application of the algorithm - constructing a DEA model with negative values in the decision-making unit and solving the problem that the traditional DEA model can not handle negative values in the decision-making unit Insufficient.