A Hybrid Multi-Attribute Decision-Making Approach Considering Correlations Among Indicators

Abstract

Considering the fuzzy and uncertainty of the evaluation information of the evaluation object, the type of indicators are expanded into five types of mixed evaluation information, namely, exact number, interval number, triangular fuzzy number, hesitant fuzzy number, and probabilistic linguistic term sets, and different types of evaluation information are adopted according to the different characteristics and types of indicators, respectively. When the evaluation information is hybrid information, the generalized Shapley function based on fuzzy measurement is used to analyze the interaction between indicators and determine the weights of indicators, considering that the interaction between indicators is more complicated. In view of the fact that the conversion of mixed information into the same kind of information will lead to the problem of complicated calculation and missing information, the VIKOR method is used to comprehensively evaluate the evaluation objects and select the best ones. Finally, the validity and feasibility of the proposed method is verified by taking the assessment of data literacy level and competence of teachers in a university in a western province of China as an example.