Designing a Hybrid Model of Data Envelopment Analysis with Taguchi Approach to Optimize Multiple Response Banks Performance

Many organizations (including banks) have a multi-step process and their operations are a continuous process in successive periods. The Taguchi method is an efficient way to optimize a single quality response. However, in practice most products / processes have more than one qualitative response. Recently, the multi-answer question in the Taguchi method has attracted considerable research attention. Therefore, this study presents an efficient approach to multi-response problem solving in Taguchi method using hybrid data envelopment analysis (DEA) model. Each experiment is discussed in the Taguchi Orthogonal Array (OA) as a decision unit (DMU) with multiple input and / or output response sets. Each DMU is evaluated by a hybrid model. The sequential DUM productivity value is then used to decide the optimal factor levels for the multi-response problem. The computational results showed that the proposed approach provides the most anticipated improvement in PCA, DEA (DEAR) and other available techniques. The suggestion may be of great help to managers in solving multi-response problems in production programs in the Taguchi method. Many organizations (including banks) have a multi-step process and their operations are a continuous process in successive periods. The Taguchi method is an efficient way to optimize a single quality response. However, in practice most products / processes have more than one qualitative response. Recently, the multi-answer question in the Taguchi method has attracted considerable research attention. Therefore, this study presents an efficient approach to multi-response problem solving in Taguchi method using hybrid data envelopment analysis (DEA) model. Each experiment is discussed in the Taguchi Orthogonal Array (OA) as a decision unit (DMU) with multiple input and / or output response sets. Each DMU is evaluated by a hybrid model. The sequential DUM productivity value is then used to decide the optimal factor levels for the multi-response problem. The computational results showed that the proposed approach provides the most anticipated improvement in PCA, DEA (DEAR) and other available techniques. The suggestion may be of great help to managers in solving multi-response problems in production programs in the Taguchi method.