Nonlinear genetic-based model for supplier selection: a comparative study

Abstract

Evaluation and selection of candidate suppliers has become a major decision in business activities around the world. In this paper, a new hybrid approach based on integration of Gene Expression Programming (GEP) with Data Envelopment Analysis (DEA) (DEA-GEP) is presented to overcome the supplier selection problem. First, suppliers’ efficiencies are obtained through applying DEA. Then, the suppliers’ related data are utilized to train GEP to find the best trained DEA-GEP algorithm for predicting efficiency score of Decision Making Units (DMUs). The aforementioned data is also used to train Artificial Neural Network (ANN) to predict efficiency scores of DMUs. The proposed hybrid DEA-GEP is compared to integrated approach of Artificial Neural Network with Data Envelopment Analysis (DEA-ANN) to support the validity of the proposed model. The obtained results clearly show that the model based on GEP not only is more accurate than the DEA-ANN model, but also presents a mathematical function for efficiency score based on input and output data set. Finally, a real-life supplier selection problem is presented to show the applicability of the proposed hybrid DEA-GEP model.