EDM Subgenre Classification Using Supervised Machine Learning Classification

The goal of this project was to use supervised learning techniques to build a model capable of classifying Electronic Dance Music (EDM) songs into their appropriate subgenres. I used the Spotipy library to access the Spotify API in order to collect 13 features for approximately 35,000 individual EDM tracks. I tested several classification techniques and selected a Random Forest model for its balance of classification accuracy and relative simplicity versus the top performing XGBoost Classifier model. The model was then fed to a Streamlit app I dubbed Genrify, which includes various tools that allow the user to classify Spotify recommendations, tracks within a playlist, and individual tracks into seven different EDM subgenres: Drum’n’Bass, Deep House, Dubstep, Hardstyle, Progressive House, Techno, and Trance.