In this funny project, conducted in collaboration with T. Granger, we tested various algorithms and machine learning methods to identify the musical genre of artists and their album inspirations. We used the Spotify API to gather data on artists from four distinct musical genres: rock, hip-hop, electronic, and classical music.

We explored machine learning techniques and data visualization methods, such as Principal Component Analysis (PCA), K-Nearest Neighbors (KNN), Bayesian Inference, and Neural Networks. One of the highlights of our study was the successful application of the Random Forest algorithm for this type of project.

Our project demonstrated that even with relatively simple algorithms, we could achieve satisfactory results. The Spotify API also proved to be a practical tool for collecting and analyzing musical data. Overall, this project showcases the power and flexibility of machine learning techniques for music analysis.

You can find the project report here (in french) and the project code on GitHub (coming soon).