Fronteiras DEA Difusas

Abstract

The \fuzzy DEA frontier is applied in situations where some variables (inputs or outputs) present a certain degree of uncertainty in its measurement, without any assumption about the probability distribution function. we build the ecien t frontier taking in account the minor and major values of the input (or output). The ecien t frontier is, in that case, a fuzzy set, to which the DMUs have a degree of membership. In order to compute the degree of membership two frontiers are constructed: the optimistic and the pessimistic frontiers. As this indicator isn’t an eciency score, we introduce the inverted frontier concept, which allows the fuzzy-DEA eciency score computation. We developed the case where only one variable present a certain degree of uncertainty and it is generalised to the case where some or all variables are uncertain.