Krishnan Anath Rau, Hamid Mohamad Rizal, Tanakinjal Geoffrey Harvey, Asli Mohammad Fadhli, Boniface Bonaventure, Ghazali Mohd Fahmi
Labuan Faculty of International Finance, University Malaysia Sabah, Jalan Sg. Pagar, 87000, Labuan F.T., Malaysia.
Faculty of Computing and Informatics, University Malaysia Sabah, Jalan Sg. Pagar, 87000, Labuan F.T., Malaysia.
MethodsX. 2023 May 30;10:102227. doi: 10.1016/j.mex.2023.102227. eCollection 2023.
Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) is a popular multi-criteria decision-making method that ranks the available alternatives by examining the ideal-positive and ideal-negative solutions for each decision criterion. The first step of using TOPSIS is to normalize the presence of incommensurable data in the decision matrix. There are several normalization methods, and the choice of these methods does affect TOPSIS results. As such, some efforts were made in the past to compare and recommend suitable normalization methods for TOPSIS. However, such studies merely compared a limited collection of normalization methods or used a noncomprehensive procedure to evaluate each method's suitability, leading to equivocal recommendations. This study, therefore, employed an alternate, comprehensive procedure to evaluate and recommend suitable benefit/cost criteria-based normalization methods for TOPSIS (out of ten methods extracted from past literature). The procedure was devised based on three evaluation metrics: the average Spearman's rank correlation, average Pearson correlation, and standard deviation metrics, combined with the Borda count technique.•The first study examined the suitability of ten benefit/cost criteria-based normalization methods over TOPSIS.•Users should combine the sum-based method and vector method into the TOPSIS application for safer decision-making.•The maximum method (version I) or Jüttler's-Körth's method has an identical effect on TOPSIS results.
逼近理想解排序法（TOPSIS）是一种流行的多标准决策方法，它通过检查每个决策标准的理想正解和理想负解来对可用备选方案进行排序。使用TOPSIS的第一步是对决策矩阵中不可通约的数据进行归一化处理。有几种归一化方法，这些方法的选择确实会影响TOPSIS的结果。因此，过去曾有人努力比较并推荐适合TOPSIS的归一化方法。然而，此类研究仅比较了有限的一组归一化方法，或使用了不全面的程序来评估每种方法的适用性，导致推荐结果模棱两可。因此，本研究采用了一种替代的、全面的程序来评估并推荐适合TOPSIS的基于效益/成本标准的归一化方法（从过去的文献中提取的十种方法）。该程序是基于三个评估指标设计的：平均斯皮尔曼等级相关性、平均皮尔逊相关性和标准差指标，并结合了博尔达计数技术。•第一项研究考察了十种基于效益/成本标准的归一化方法对TOPSIS的适用性。•用户应将基于总和的方法和向量方法结合到TOPSIS应用中，以进行更安全的决策。•最大方法（版本I）或尤特勒-科尔特方法对TOPSIS结果有相同的影响。
