Support vector classifiers ensemble based on fuzzy integral

For the constraint of traditional voting-based support vector classifiers (SVCs) ensemble technique to the classification performance due to the impossibility of evaluating the importance degree of the output of individual component of SVC to the final decision, an SVCs ensemble method based on fuzzy integral is proposed. This method considers not only the objective information for the outputs of each component of SVC, but also the importance degree of the output of individual component of SVC to the final decision. Therefore, the classification performance is enhanced to a great extent. Simulation results demonstrate that the proposed SVCs ensemble approach based on fuzzy integral outperforms a single SVC and traditional SVCs ensemble technique via majority voting.