Solving multi-response optimization problem using artificial neural network and PCR-VIKOR

Abstract

In this paper a hybrid approach is introduced to solve multiple response problems. In the proposed method signal to noise (SN) ratio is computed and then SN ratios for unexperimented treatments are estimated using artificial neural network. The SN ratios are converted into a process performance index by applying process capability ratio and VIKOR method, so the treatments can be ranked and the best of them is selected. The performance of the proposed method is verified in a case study. Moreover a sensitivity analysis has been done by a VIKOR score estimator turned neural network. The results show efficiency of the proposed approach.