A simulation-based DEA approach for multiple criteria decision-making problems with uncertain mixed-criteria values

¹ College of Mathematics and Informatics, South China Agricultural University, 483 Wushan Road, Guangzhou 510000, P.R. China
² International School of Business Finance, Sun Yat-Sen University, Zhuhai Campus, Tangjiawan, Zhuhai 519082, P.R. China
³ School of Management, Hefei University of Technology, 193 Tunxi Road, Hefei 230009, P.R. China
⁴ Intelligent Interconnection System Anhui Provincial Laboratory, Hefei 230009, P.R. China
⁵ School of Management, University of Science and Technology of China, 96 Jinzhai Road, Hefei 230026, P.R. China

^* Corresponding author: shengang@hfut.edu.cn

Received: 21 June 2022
Accepted: 14 May 2024

Abstract

In ex-ante decision scenarios, predicting criterion values accurately is difficult for decision makers (DMs). Inconsiderable work is normally required for measuring criteria by uncertain random values or ordinal values. However, in the classical data envelopment analysis (DEA) model, criterion values are the constants that limit the application of the classical DEA model in ex-ante decision scenarios. This paper presents a simulation-based DEA approach, which captures random and ordinal criterion values by a simple and direct simulation-based approach. The approach includes three steps. In the first step, Monte Carlo simulation methods are used to convert uncertain random values or ordinal values into cardinal data. In the second step, we use traditional DEA methods to compute the efficiency score of decision-making units (DMUs). In the third step, we ranked all DMUs by calculating the DEA-efficient acceptability of each DMU in multiple simulations and then selected the optimal DMU. The proposed approach is illustrated by experimental examples and a case study of a municipal wastewater treatment system.