Comparative Evaluation of Various Models for the Classification of Music Genre

This project explores music genre differentiation using the GT-ZAN dataset, comprising ten genres. Ten classical machine learning models, including KNN, SVM, and Cross Gradient Booster, are compared using features such as MFCC mean, MFCC variance, genre labels, and root mean square values from the frequency and time domains. Experimental results demonstrate variations in model performance, with specific classifiers achieving higher accuracy and precision for particular genres, while others excel in recall and F1-score. The findings offer valuable insights for future music genre classification and contribute to the development of music genre categorization systems.