DEA-DA Model and Method with L-R Fuzzy Numbers

DEA-DA was a new non-parametric method which incorporated a methodological strength of DEA into the discriminant analysis formulation. To overcome the shortcoming of DEA-DA which could only deal with crisp observational data, a DEA-DA model using L-R fuzzy numbers as observational datum, denoted as FDEA-DA, was proposed in this article. The FDEA-DA model could dispose L-R fuzzy records and simultaneously maintain its discriminant capability in fuzzy control. Two examples were selected in this article to help to illustrate the disparity of classification when original crisp data was changed into corresponding L-R fuzzy numbers. And it could identify the FDEA-DA method´s capability of dealing with multidimensional fuzzy data