Pengklasifikasian Genre Musik Berdasarkan Sinyal Audio Menggunakan Support Vector Machine

Abstract

Musical genre is a label for musical art to characterize and categorize music. The determination of musical genre is done based on the similarity between the music. Most of the classification of musical genre is done manually, which requires a lot of effort and time especially when there is a big database of music. The purpose of this research is to build a support vector machine model to automatically classify the musical genre and implement the method to extract the musical surface and rhythm features. The developed model is then utilized to determine the genre from the unknown music. Support vector machine is a learning system whose classification uses a hypothesis space in the form of linear functions in a high dimension of feature space. The extracted musical surface and rhythm features are the centroid, roll off, flux, zero crossings, low energy, and four peaks of the signal‟s autocorrelation calculation. The result of this research is a classification model with an average accuracy of 65%.