The Research Of The Management Method Based On AFS Theory And Its Application

Forecasting and decision-making are the two important parts of management science. In recent years, they have been studied by many scholars. The developed methods have been successfully applied to the actual field of engineering, economic, and market analysis. This paper analyzes and improves the representative prediction method-time series analysis and the most commonly used method of decision-making-multi-attribute decision making method, by the application of Axiomatic Fuzzy Set (AFS) fuzzy theory. The main contributions are the followings.(1) Firstly, a multiplicative seasonal model was established for container throughput of Shanghai Port. Secondly, proposing a fuzzy time series forecasting method based on AFS theory, which predict the fuzzy time series according to the fuzzy trend of AFS membership:(ⅰ) the AFS fuzzy membership function is applied to calculate each fuzzy set membership on the historical data to divide the domain;(ⅱ) fuzzified the raw data according to AFS fuzzy logic so as to establish the fuzzy relations;(ⅲ) forecasting the historical data according to the trend of fuzzy membership.(2) Firstly, presenting a Integrated multi-attribute decision making method based on AFS theory:(ⅰ) selecting a best combination from the initial data set by the application of data envelopment analysis (DEA);(ⅱ) calculating the weight values of attributes by combining the AFS fuzzy description algorithm and Analytic Hierarchy Process (AHP);(ⅲ) applying the technique for order preference by similarity to ideal solution(TOPSIS) to make a final sort in the best combination. Secondly, applying AFS theory to extend TOPSIS method in fuzzy environment, which mainly transform the candidate unit contained the semantic variable or the vague concept into its membership value in the fuzzy concept by AFS fuzzy membership function and then carrying out the steps of TOPSIS method on its membership degree. This fuzzy method can deal with various types of data including semantic variables, fuzzy concept and Boolean data so that the application range of TOPSIS method has been greatly expanded. In addition, a clear semantic interpretation of the decision-making result was given according to the AFS fuzzy description. (3) Designing a online banking performance evaluation method based by on the DEA and AFS clustering algorithm. First of all, deleting part of the less efficient banks from all the online banking through the DEA method; next, applying AFS clustering algorithm to cluster the remaining banks, and giving the semantic description of each category.Finally, the thesis draws the conclusion on the researches and discusses about the further study.