This is a repository for a music genre classification project using traditional machine learning models and deep learning.
The dataset used in this project is the Free Music Archive (FMA) dataset.
Dataset link: https://github.com/mdeff/fma
datapreprocessing.py: data preprocessing and feature extraction.
traditional_ml.ipynb: music genre classification using traditional ML methods such as random forests, k-nearest neighbours and support vector machine. It also includes a majority vote classifier that combines the predicitions from the previously mentioned methods
deep_learning.ipynb: music genre classification using deep learning including three different archiectures.
feature_illustration.ipynb: illustrations of the features extracted in this project.
