DEA数据标准化方法及其在方向距离函数模型中的应用

Directional distance function is the generalization of radial model in data envelopment analysis. It has the capacity of dealing with undesirable outputs, but the problem is that it has no unit-invariant measurement of efficiency, which hampers its application to empirical studies. Data normalization for data envelopment analysis is a universal solution for the problem of unit-invariance, and the efficiency keeps unchanged in radial and non-radial models after data normalization. A unit-invariant efficiency measure for directional distance function is developed based on DEA data normalization.