Music Genre Classification Using Hybrid Committees and Voting Mechanisms

In this paper, using the example of music genre classification, seven voting methods for classifier committees are presented and used. This work focuses on improving the results of classification quality mainly by using different voting methods in different classifier committees. Examination of these methods was carried out on four classifier committees created from a total of 31 individual models. Classification quality is crucial in many machine learning applications. Therefore, methods for improving this quality may be of great practical value. The use of classifier committees and the proposed voting methods significantly improved the results comparing to base models.