Improving the commercial success of music on streaming platforms using data science

This paper proposes a machine learning model that can predict whether a song is likely to crack the top 50 songs globally on music streaming platforms such as Spotify, Apple Music, etc. Backed by the dataset of the top songs across the years it tracks parameters like acousticness, danceability, energiness, genre, etc to identify the characteristics of the ideal songs. We propose an SVM based model and a logistic regression based model to predict whether an input song will be popular. We also propose the ideal values of the aforementioned attributes which are common across songs in the top 50 as observed in the previous years in our dataset built using music information retrieval systems.