An output‐oriented classification of multiple attribute decision‐making techniques based on fuzzy c‐means clustering method

Abstract

This paper presents a new output‐oriented classification of multiple attribute decision‐making (MADM) techniques, not based on common subjective comparisons, but mostly on quantitative and computer‐aided comparisons and results. Several classifications of MADM techniques exist, all of which are either input‐oriented (based on the type of input data) or process‐oriented (depending on the process of calculating the final results using the input data). The classification provided in this paper is based on measuring the performance of 17 MADM techniques (SAW, ELECTRE I, TOPSIS, ORESTE, PROMETHEE I, EVAMIX, MAUT, REGIME, MAPPAC, TACTIC, VIKOR, ARGUS, COPRAS, SMART, PACMAN, MOORA, and ARAS) in seven performance variables (simplicity in learning and deploying, speed, complexity of calculations, the number of inputs, the quality of the underlying logic, the quality of rankings, and the rate of growth in large problems) and clustering them using fuzzy c‐means clustering method. Results indicate that the considered techniques can be best classified into two clusters.

Food market segmentation based on consumer preferences using outranking multicriteria approaches

“…This is an important difference between these two approaches. In addition, segments from cluster analysis can overlap (Asgharizadeh et al., 2019) and lack stability (Van der Zanden et al., 2014; Hajibaba and Grün, 2020). In the literature, the number of segments in the food domain range from four to five, with one or two careless or uninvolved segments (Verain et al., 2012), which can be compared to S4 groups in the multicriteria approach of this research.…”

“…In addition, the multivariate statistical analysis applied in marketing has mainly consisted of clustering algorithms for searching consumer segments and discriminant analysis in order to characterize them. Some authors highlight cluster overlapping (Asgharizadeh et al., 2019) from real data and the difficulty of obtaining compact groups of customers. The lack of robustness and stability of generated clusters (Wang et al., 2008; Van der Zanden et al., 2014; Hajibaba and Grün, 2020), as well as the difficulty managers face in characterizing and understanding them, are common criticisms (Yankelovich and Meer, 2006).…”

Assessment decisions of independent learning activities using SMART–FCM method

“…The results of these studies indicate that the SMART method can produce more structured, systematic, and transparent results with a success rate. In 2017, Asgharizadeh et al [13] discussed the new output-oriented classification of the Multiple Attribute Decision Making (MADM) techniques, where the classification was input-oriented or process-oriented. In measuring the performance of seventeen MADM techniques for classification results based on seven performance variables, the results of each MADM technique would be clustered using FCM.…”

The provably good parallel seeding algorithms for the k‐means problem with penalties

“…One can choose k cluster centers randomly and cluster the original points to these centers, then update the clustering by computing the center of mass of these clusters, and so on. This method is also applied to solve the variants of k-means problem (Asgharizadeh et al, 2019), while there are many examples showing that the approximate ratio cannot be guaranteed. In order to improve the accuracy, the seeding algorithm, whose task is to find the initial cluster center, is given and then is improved by Aggarwal et al (2009), Bachem et al (2016bBachem et al ( , 2017, and Ostrovsky et al (2013).…”