An improved combination method of multiple classifiers based on fuzzy integrals

Combination of multiple classifiers is a powerful solution to difficult pattern recognition problems. Fuzzy integral is one of the many effective ways to combine classifiers, but how to calculate the fuzzy integral density with fuzzy integrals remains an unsolved problem. A combination method based on the fuzzy integral and Bayes method is presented to obtain the fuzzy integral density function, and then combines the output information of all classifiers with the fuzzy integral. Experiments show that this method achieves a very promising performance and outperforms other combining approaches.