Why does a song like 'Shape of You' or 'Sorry' lead the charts? What determines whether a song is going to be a hit or a flop? What factors determine the popularity of a song?
I tried to answer these questions using my Machine Learning knowledge. Many feel the popularity of a song comes from the popularity of the artist. Only well-known artists make it to the radio / top charts. Social media, radio, and award shows force us to listen to these songs.Â
But, what if we take the popularity of an artist out of the picture? Does a catchy rhyme or a good rhythm affect the popularity of a song? To answer this, I tried to predict the popularity of a song using discernible features like its energy, loudness, danceability, instrumentalness etc.
I also used this opportunity to test the prediction accuracy of different Machine Learning models and compare the results. The models used for this project were: Multiple Linear Regression, Polynomial Regression, SVM Regression, Decision Tree Regression, Random Forest Regression.