The inability to curate desirable music also causes audiences move between platforms searching for music they enjoy ( Prey, 2018). The inability to predict hits means that artists are often underpaid for their work and music labels misallocate production and marketing resources when seeking to build audiences for new music ( Byun, 2016). This has been called the “Hit Song Science” problem ( McFee et al., 2012). Unfortunately, the accuracy of predictions has generally been low ( Prey, 2018). Music distribution channels use both human listeners and artificial intelligence models to identify new music that is likely to become a hit. The surfeit of choices makes it difficult for streaming services and radio stations to identify songs to add to playlists.

Our results demonstrate that applying machine learning to neural data can substantially increase classification accuracy for difficult to predict market outcomes.Įvery day, 24,000 new songs are released worldwide ( Pandora, 2018). Applying machine learning to the neural response to 1st min of songs accurately classified hits 82% of the time showing that the brain rapidly identifies hit music. This model classified hit songs with 97% accuracy. Then, we created a synthetic set data and applied ensemble machine learning to capture inherent non-linearities in neural data. A linear statistical model using two neural measures identified hits with 69% accuracy. We compared several statistical approaches to examine the predictive accuracy of each technique. We took a different methodological approach, measuring neurophysiologic responses to a set of songs provided by a streaming music service that identified hits and flops. Traditionally, song elements have been measured from large databases to identify the lyrical aspects of hits. Identifying hit songs is notoriously difficult. 2Immersion Neuroscience, Henderson, NV, United States.1Center for Neuroeconomics Studies, Claremont Graduate University, Claremont, CA, United States.

Are spotify lyrics accurate