Using Genetic Algorithms to Identify Fuzzy Measures for Multiple Attribute Decision Making

General fuzzy measures that fulfill only the boundary conditions and monotonicity can be employed to analyze multiple attribute decision making problems. Moreover, people's subjective evaluation processes can be approximated more accurately using fuzzy integrals. However, fuzzy integrals are not differentiable with respect to fuzzy measures. In addition, the problem size is overtly large and there exists rather complicated interaction among the attributes. Thus, the identification of general fuzzy measures is insurmountable. This study develops an identification procedure for general fuzzy measures using genetic algorithms and conducts an experimental analysis toward the simulated data. The satisfactory results indicate that the fuzzy measure identification method by genetic algorithms performs well.