Genetic algorithm for integration of ensemble of classifiers in arrhythmia recognition

The paper presents application of the genetic algorithm for the integration of classifiers used in the ensemble for the accurate recognition of arrhythmia on the basis of ECG waveform. It is well known that using many classifiers arranged in the form of ensemble leads to the increased accuracy of recognition. However the important problem is the integration of them into one classification system. The paper shows that application of the genetic algorithm allows to reduce significantly the total error of recognition. The comparison to other known methods of integration proves the superiority of the proposed approach.