IAM Profect Ancri Carlo.

This project focuses on automatic music genre classification using machine learning techniques. Audio samples from three genres—Electronic, Jazz, and Metal—were processed to extract relevant features such as MFCCs and chroma vectors. Various classifiers, including Decision Trees (DT) and k-Nearest Neighbors (kNN), were trained and evaluated on both clean and noisy datasets. The results show how different feature sets and noise levels impact classification accuracy, highlighting the challenges of genre recognition, especially with limited data and under real-world audio conditions.