Open AccessStatistical Analysis of the Reliability of a Decision Obtained by the TOPSIS Method
Author(s) -
Rūta Simanavičienė,
Jovita Cibulskaitė
Publication year - 2019
Publication title -
lithuanian journal of statistics
Language(s) - English
Resource type - Journals
ISSN - 2029-7262
DOI - 10.15388/ljs.2015.13886
Subject(s) - topsis , reliability (semiconductor) , computer science , measure (data warehouse) , decision analysis , statistical analysis , multiple criteria decision analysis , operations research , statistics , reliability engineering , data mining , mathematics , engineering , power (physics) , physics , quantum mechanics
The tasks of making the most appropriate decisions taking into account a number of criteria are dealt with in variousfields such as engineering, industry, finance, economics and others. If the aim is to arrange the alternatives in a priority lineaccording to quantitative attributes, then multiattribute decision-making methods are suitable. Analysts using these methods usuallydo not take into account initial data errors – deviations in attribute values, in which case the decision may be unreliable. In thisarticle, several statistical analysis methods are proposed for the multicriteria decision to measure reliability: formulation of statisticalhypotheses and calculation of confidence intervals for parameters. Based on statistical analysis results, conclusions about thereliability of a multicriteria decision obtained using the TOPSIS method are formulated.