Klasifikasi Genre Musik Menggunakan Support Vector Machine Berdasarkan Spectral Features

Abstract

This research focuses on music genre classification based on spectral features and SupportVector Machine (SVM). Features such as Spectral Centroid, Spectral Rolloff, Spectral Flux, and Spectral Bandwidth are extracted from MP3 music audio. The dataset comprising 4 music genres is utilized for training and testing the system. The extracted spectral features are fed into the SVM classifier to predict the genre of test samples. Python and machine learning are both used in developing the system while the experimental results demonstrate the effectiveness of SVM in accurately classifying music genres based on the current extracted features. The proposed approach contributes to automated music genre classification systems, facilitating music organization, recommendation, and retrieval. This research promotes advancements in music information retrieval and enhances user experience in music-related applications.