Researchers developed a machine learning model that uses neural responses to predict the success of songs with 97% accuracy. In their study, participants listened to a set of songs while their neurophysiologic responses were monitored, generating data that helped the machine learning model determine potential hits.
Streaming services and radio stations face the challenge of choosing songs for playlists from the overwhelming daily release of tens of thousands of tracks. To overcome this, a US study utilized machine learning and brain responses to predict hit songs with an impressive 97% accuracy. By analyzing neurophysiologic data, researchers identified hits with unprecedented precision. This approach, known as ‘neuroforecasting,’ enables accurate predictions based on a small group’s neural activity. The findings offer streaming services the ability to efficiently identify popular songs, enhancing user experience. Future applications may extend beyond music, benefiting movie and TV show recommendations.
Source NeuroScienceNews
