Year: 2020
Type: conference-paper
Abstract: Multi-criteria decision-methods (MCDA) has a broad field of applications in various areas, such as engineering, logistic, health management, and others. Their main objective is to rank the decision variants, from the best to the worst. It is always essential to get a reliable solution to a decision problem. However, rankings obtained with different MCDA methods are often different, and that is why an interesting research gap is related to measuring the accuracy and reliability of MCDA methods. In this paper, we examine how much different can be rankings obtained using TOPSIS and VIKOR methods. For this purpose, we apply classical versions of TOPSIS and VIKOR methods to randomly generated decision matrices with different criteria and alternatives. Then, we compare the obtained ranking using three ranking similarity coefficients. The comparison results of conducted simulations are represented as boxplots, which contain the primary distribution of similarity of results calculated using the tested methods. The work is finished with a discussion and conclusions on the results in different classes of problems.
Source: Procedia Computer Science
Authors: Andrii Shekhovtsov, Wojciech Sałabun
Institution: West Pomeranian University of Technology in Szczecin
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: diamond
Funder: Narodowe Centrum Nauki
OpenAlex ID: W3091016193