Perusing the influence of desired value on Back Propagation Neural Network using Data Envelopment Analysis Models in Supplier Selection Evaluation

Based on the complexity and diversity of the features of the supplier efficiency and performanceevaluation toady, neural networks can be used. There are several papers that use supervisedneural networks to evaluate the efficiency of the decision making units, by a set of desired valueneeded.This paper aims to peruse the quality of the desired value on the neutral network training andpredicting. Hence here we use SBM, SBM-Output oriented and Russell models efficiency scoresto introduce new sets of desired value in supplier selection evaluation process by neural network.Then compare the gained results together to find the best set of desired value for network forsupplier selection evaluation.

کلیدواژه ها:

Data Envelopment Analysis (DEA) ، Back Propagation Neural Network (BPNN) ، Desire Value