250 / 2024-04-23 20:54:24

TOPSIS-Sort-T: A TOPSIS-Sort Method Based on Tolerance Interval Estimation

TOPSIS-Sort,TOPSIS,Ranking Reversal,Tolerance Interval Estimation

In this study, a multi-criteria decision sorting (MCDS) method based on tolerance interval estimation, TOPSIS-Sort-T, is proposed, which does not need to provide boundary profiles information and is able to prevent the influence of the rank reversal problem existing in TOPSIS on the sorting results as well. Firstly, the current set of alternatives to be evaluated is regarded as a sample set of alternatives drawn from the overall set of alternatives, and the tolerance interval estimation method is used to estimate the maximum and minimum possible evaluation values under each criteria of the overall set of alternatives, which are added to the current set of alternatives to be evaluated as virtual alternatives. Secondly, an expanded probability grading method is used to determine the proportion of the range of evaluation values under each attribute for each classification, and then the boundaries profiles are determined based on the tolerance interval estimation method and added to the set of alternatives to be evaluated. Thirdly, the relative nearness degree of each alternative, virtual alternatives, and boundary profiles is calculated based on TOPSIS, and the alternatives are classified according to the evaluation results. Finally, the proposed method is compared with other MCDS methods to verify its effectiveness and superiority.