This project was build for the course named Machine Learning at the School of Engineering and Applied Science (SEAS), Ahmedabad University (AU)
Spotify is one of the leading audio streaming platforms, hosting almost all song collections present in the world. We selected an independent dataset derived from Spotify Public API comprising of more than 1.2 Million songs, to design a machine learning model for the song recommendation system. The recommendation is done based on different audio features by the means unsupervised learning algorithms which classifies the data into different clusters. The system will recommend songs belonging to a particular set of cluster based on different features of the song. The dataset consists of various attributes and features of the songs, which helps in grouping them based on certain similar premises.
The implementation can be seen here and the following results can be see from here.