Zakeri Shervin, Konstantas Dimitri, Chatterjee Prasenjit, Zavadskas Edmundas Kazimieras
Geneva School of Economics and Management, University of Geneva, 1211, Geneva, Switzerland.
Faculty of Civil Engineering, Institute of Sustainable Construction, Laboratory of Smart Building Systems, Vilnius Gediminas Technical University, Vilnius, Lithuania.
Sci Rep. 2025 Jan 2;15(1):284. doi: 10.1038/s41598-024-81027-4.
Subjective weighting methods are widely employed to determine criteria weights in multi-criteria decision-making (MCDM) environment. Inputs from decision-makers, including opinions, assessments, assumptions, evaluations, interpretations, expectations, and judgments, are primarily relied upon in these methods. Significant challenges are faced due to two primary factors: the inherent uncertainty in inputs and the process of pairwise comparisons. These challenges increase the uncertainty regarding the derived weights, raising concerns about the reliability of such approaches. This paper introduces a novel MCDM method called Soft Clusters-Rectangles (SCR) to overcome such limitations. This method distinguishes itself by avoiding pairwise comparisons and adopting a fuzzy approach to address uncertainty. Weights are calculated based on criteria membership values across three defined clusters, namely immaterial, mediocre, and vital. Each cluster represents a distinct range of importance. Analytic geometry is also used by computing the areas of multiple rectangles to determine the final weights. The application of SCR method is demonstrated through a case study on autonomous vehicle route selection problem. The results are thoroughly examined through comprehensive analysis. The findings reveal that this method not only avoids the challenges posed by pairwise comparison-based methodologies but also exhibits similarities with objective weighting techniques, often yielding results more comparable to those produced by hybrid methods.
主观加权方法在多准则决策（MCDM）环境中被广泛用于确定准则权重。这些方法主要依赖决策者的输入，包括意见、评估、假设、评价、解释、期望和判断。由于两个主要因素，面临着重大挑战：输入中的固有不确定性和成对比较过程。这些挑战增加了导出权重的不确定性，引发了对此类方法可靠性的担忧。本文介绍了一种名为软聚类 - 矩形（SCR）的新型MCDM方法来克服这些限制。该方法的独特之处在于避免成对比较，并采用模糊方法来解决不确定性。权重是基于跨越三个定义的聚类（即无关紧要、中等重要和至关重要）的准则隶属值来计算的。每个聚类代表不同的重要性范围。还通过计算多个矩形的面积来使用解析几何确定最终权重。通过一个关于自动驾驶车辆路线选择问题的案例研究展示了SCR方法的应用。通过全面分析对结果进行了深入研究。结果表明，该方法不仅避免了基于成对比较的方法所带来的挑战，而且与客观加权技术有相似之处，其结果通常与混合方法产生的结果更具可比性。
