Abstract: In order to solve the output-oriented Data Envelopment Analysis（DEA） model in fuzzy environment, based on credibility theory, this paper provides the output-oriented Credibility DEA（CDEA） mode, in which the concept of fuzzy chance-constrained programming is adopted on objective function and all constraints have fuzzy inputs and fuzzy outputs. In some special cases, such as the fuzzy inputs and fuzzy outputs are independent trapezoidal fuzzy variables, this kind of CDEA model can be changed into the crisp equivalent class. Furthermore, two basic properties about the proposed CDEA model are discussed. At last, a numerical example is given to illustrate the effectiveness of the proposed CDEA model.

Key words: credibility theory, Data Envelopment Analysis（DEA）, fuzzy chance-constrained, efficiency analysis

摘要： 为求解模糊环境下输出倾向的数据包络分析（DEA）模型，利用可信性测度建立了一类新的输出倾向的可信性DEA（CDEA）模型，其中目标函数采用了模糊机会约束规划的概念，且所有的约束条件中都含有模糊输入和模糊输出数据。当模糊输入和模糊输出数据为相互独立的梯形模糊变量时，把所建立的CDEA模型转化为其清晰等价形式，进而研究了模型的两个基本性质；通过一个应用实例来说明所建立CDEA模型的有效性。

关键词: 可信性理论, 数据包络分析, 模糊机会约束, 效率分析