A DEA-RA Model with Restrain Cone for Efficiency Estimation and Evaluation

Data envelopment analysis (DEA) is a technique to measure relative efficiency. Regression analysis (RA) is a statistical process for estimating the relationships among variables based on the model of averaged images. Based on the continuity of human thinking, it is feasible to establish an intermediate evaluation model via combining DEA and RA. Besides, a cone-ratio DEA model C
²
WH can reflect decision makers’ preference for sample indexes or decision-making units (DMU). Thus, DEA-RA model is established based on restraint cone by combining C2WH model and RA model, and a generalized DEA-RA model is proposed on condition that restraint cone is a polyhedral convex cone. Then, the performances of 16 Chinese logistics companies are analyzed using generalized DEA-RA model, indicating that the model can reflect preferences of decision maker during judgment and analysis, and consequently a subjective and objective comprehensive evaluation method is formed.