DAUGIATIKSLIŲ SPRENDIMO PRIĖMIMO METODŲ JAUTRUMO ANALIZĖ TAIKANT MONTE KARLO MODELIAVIMĄ
THE SENSITIVITY ANALYSIS OF THE MULTIATTRIBUTE DECISION MAKING METHODS BY MONTE CARLO SIMULATION

Summary/Abstract: The quantitative multicriteria decision making methods are used for decision making, but the biases of the values of the attributes are ignored often. Therefore it is not clear, if the final decision were changed, when changing the values of attributes. There are scientific research papers intended for the sensitivity analysis of multicriteria decisions, according to the significances of attributes. This paper analyses the sensitivity of quantitative multiattribute decision making TOPSIS – Technique for Order Preference by Similarity to Ideal Solution, SAW – Simple Additive Weighting, COPRAS – COmplex PRoportional ASsessment methods, according to attribute distribution. On this basis of Monte Carlo modeling (simulation) the generation of pseudorandom attribute values is being made, according to uniform and normal distribution. Using generated data the efficiency of alternatives is carried out by the above methods. As a result the sensitivity of the methods and the reliability of final decision are presented. When multiattribute decision making methods are used on the modeling decision support systems it is recommended to consider the sensitivity of these methods.

• Details
• Contents