Classifying Music Genre and Recommending songs using various ML Algorithms

The task of music genre classification in the realm of Music Information Retrieval is considered very challenging because audio features have varied complexity and overlapping nature. This paper addresses the issue by employing, comparing, and experimenting on diverse machine learning and deep learning models, including XGBoost, Random Forest, K-Nearest Neighbors (KNN), Support Vector Machines (SVM), and Convolutional Neural Networks (CNN). The experiments are conducted on the GTZAN dataset, which has widely been used as a benchmark for music genre classification. Among the models implemented, the XGBoost demonstrated the best performance in predictions of the genre labels, acquiring a higher degree of accuracy as compared to the classical machine learning techniques. Besides, the study includes the development of a music recommendation system that suggests songs based on the input audio. It indicates that machine learning methods, like XGBoost, are highly effective for genre classification, but further enhancements, such as feature optimization and dataset expansion, could improve the performance of more traditional models.