Classification and Categorization of Music Genre Using Machine Learning Algorithms

Abstract

Music genre classification is crucial in music information retrieval for organizing large music libraries and enhancing recommendation systems. This research presents an innovative web-based application for music genre classification. Users can upload WAV format music files, which are processed to extract audio features such as spectral characteristics, rhythmic attributes, and Mel-Frequency Cepstral Coefficients (MFCCs) using the Librosa library. Preprocessing includes handling missing values, normalization, and dimensionality reduction via Principal Component Analysis (PCA). Machine learning classifiers—K-Nearest Neighbors (KNN), Support Vector Machine (SVM), and Random Forest—are trained on labelled music samples to predict genres accurately.

Classifier performance is evaluated using accuracy, precision, recall, and F1-score, with results visualized graphically. Experimental results demonstrate the approach's efficacy in inaccurate genre classification, highlighting its potential for improving music recommendation systems and genre-based analysis. In summary, the proposed music genre classification system represents a significant advancement in automated music categorization, providing a robust, efficient, and user-friendly solution that revolutionizes genre identification and classification.