This project is relevant to music lovers. Not only will it help to determine how accurate is the musical preference of a listener, it could also help understand which audio features are likely to become more pleasant than others. An ensemble to build machine learning algorithms that can predict the musical taste (like/dislike) of a user based on the tracks presented in his Spotify playlist will be made.
The following analysis also explores the audio features of the songs and extracted from the Spotify Web API. These available features include attributes such as: acousticness, danceability, duration_ms, energy, instrumentalness, key, liveness, loudness, mode, speechiness, tempo, time signature and valence.