Abstract

This study developed a hybrid supplier selection model, which combines multi-objective data envelopment analysis (MODEA) and multi-objective genetic algorithm (MOGA), called as MODP. MODEA involved the consideration of several supplier selection factors. In addition to maximizing the efficiency of the decision making unit (DMU) itself, the approach also took into account other DMUs, enabling minimization of the total deviation and minimization of the maximal deviation. Besides, to solve the MODEA model efficiently, this study applied non-dominated sorting genetic algorithm II (NSGA-II) to problem solving. Finally, the proposed MODP methodology was introduced in a hemadynamometer supplier selection case, and results suggested the proposed MODP methodology is effective for supplier selection.

Key words: Multi-objective data envelopment analysis, multi-objective genetic algorithm, supplier selection.

Abbreviation

Abbreviations: MODEA, multi-objective data envelopment analysis; MOGA, multi-objective genetic algorithm; DMU, decision making unit; NSGA, non-dominated sorting genetic algorithm; MCDM, multi-criterion decision making; CROs, contract research organizations; STB, smaller-the-better; BTB, bigger-the-better; ER, error ratio; DEA, data envelopment analysis; CEO, chief executive officer.