Comparison of the Methods Used in Multicriteria Decision-Making to Determine the Values of the Coefficients of Importanc
- PDF / 128,293 Bytes
- 10 Pages / 594 x 792 pts Page_size
- 40 Downloads / 186 Views
COMPARISON OF THE METHODS USED IN MULTICRITERIA DECISION-MAKING TO DETERMINE THE VALUES OF THE COEFFICIENTS OF IMPORTANCE OF INDICATORS THAT CHARACTERIZE A COMPLEX SYSTEM M. M. Potomkin,1† A. A. Sedlyar,1‡ O. V. Deineha,1†† and O. P. Kravets1‡‡
UDC 519.816
Abstract. This paper provides a general description of the following typical methods that are used in multicriteria decision-making to determine the values of the coefficients of importance of indicators that characterize a composite system: the analytic hierarchy process, the methods of critical distance, of pairwise comparison, and of rank, the Fishburn, the CRITIS, and the entropy methods. The features of these methods are determined and calculations are carried out that illustrate the differences in the values of the coefficients of importance of the obtained indicators with their use. The recommendations on the practical application of these methods are provided. Keywords: analytic hierarchy process, critical distance method, pairwise comparison method, rank method, Fishburn method, CRITIS method, coefficients of indicator importance, entropy method. INTRODUCTION It is common knowledge that multicriteria decision-making methods are widely used when investigating complex systems. The analysis results of a number of these methods, including TOPSIS [1], ELECTRE ²² [2], taxonomy [3], ARAS [4], WS [5], SAW [6], WASPAS [7], PROMETHEE [8], TODIM [9], VIKOR [10], MOORA [11], COPRAS [12], OCRA [13] methods, etc., attest to the fact that one of their integral output data types is the value of importance coefficient indicators that characterize the complex system under study. A comparison of its alternative variants is performed by using these values. To obtain the values of the indicator importance coefficients, there exists a number of methods that are fundamentally different from each other in the principal ideas they are based upon; therefore, we surmise that a comparison of the methods used during multicriteria decision-making to obtain these coefficients is a currently important scientific assignment. REVIEW OF METHODS USED TO OBTAIN THE VALUES OF INDICATOR IMPORTANCE COEFFICIENTS As is known, a typical decision-making problem statement where the above-mentioned methods are used is as follows. Let there be the number n of alternative variants of a certain system, each of them characterized by the indicator value m upon which the systems performance is assessed. These values are specified by a matrix with the elements E ij (i = 1, K , n ; 1
Central Scientific and Research Institute of the Armed Forces of Ukraine, Kyiv, Ukraine, †[email protected]; [email protected]; ††[email protected]; ‡‡[email protected]. Translated from Kibernetika i Sistemnyi Analiz, No. 6, November– December, 2020, pp. 149–159. Original article submitted June 10, 2019.
‡
990
1060-0396/20/5606-0990 ©2020 Springer Science+Business Media, LLC
j = 1, K , m). Furthermore, there are m importance coefficients whose values characterize the importance of each of the indicators with consider
Data Loading...
