Open AccessOn ranking by using weighted self-normalizing distance metrics in multi-attribute decision-making
Author(s) -
M. Souissi,
Sana Hafdhi
Publication year - 2021
Publication title -
decision science letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.384
H-Index - 18
eISSN - 1929-5804
pISSN - 1929-5812
DOI - 10.5267/j.dsl.2021.7.003
Subject(s) - normalization (sociology) , ranking (information retrieval) , computer science , data mining , multiple criteria decision analysis , selection (genetic algorithm) , group decision making , operations research , artificial intelligence , mathematics , sociology , anthropology , political science , law
Preliminary normalization is central to the decision process of several popular, recent or completely new multi-attribute decision-making (MADM) methods. However, a number of authors have pointed out serious pitfalls attributed to normalization methods. One major pitfall, which has been identified, is that normalization methods may lead to different final rankings of alternatives when a ranking procedure (RP) based on them is used for solving a MADM problem. The current paper aims to ascertain and illustrate the effectiveness of some RPs based on prominent primary WEighted Self-NORmalizing Distance (WESNORD) metrics and their averages. The effectiveness of the selected RPs is demonstrated by solving a logistics service provider (LSP) selection problem taken from the literature. The results reveal that the RPs considered deliver final rankings of alternatives, which are very similar to the SAW-produced reference ranking.