Research On Chance-constrained DEA Models With Fuzzy Factors

This paper puts forward fuzzy chance-constrained N programming based on fuzzy chance-constrained P programming and necessity measure, and makes an extension of the framework of fuzzy chance-constrained programming. According to management and behavior theory, the concept of base efficiency is established. Referring to P, N programming, two fuzzy chance constrained DEA models based on P, N measure, FCCPDEA, FCCNDEA are proposed, which handle fuzzy uncertainty with confidence level and identify the efficient decision making units with efficiency discriminator, and reach the conclusion that at least one DMU (decision making unit) is efficient relative to its base efficiency with a certain confidence level.Possible production sets and production frontiers under fuzzy chance constraints are constructed, and fixed-figure projection analyses are made with the data under chance constraints. Similar coefficients are defined to make clustering analysis without calculation of DMU's efficiency by the difference between efficiency discriminator and the confidence level.FCCPDEA, FCCNDEA reach similar theorems and conclusions, but they have different theoretical and practical meaning from the aspects of theoretical framework,objective function, factors affecting efficiency discriminators, possible production frontiers.FCCPDEA and FCCNDEA, which are non-linear programming models combined with human value judgment, can handle problems with fuzzy impression. Under some conditions, they are more general extension of DEA methodology. They can also make evaluation of perceived service quality.