Soundtrack Success: Unveiling Song Popularity Patterns Using Machine Learning Implementation

Music streaming services are getting increasingly popular as a result of the prevalent need for web and smart gadgets. Melophiles are drawn toward a range of musical genres and create a unique digital footprint. The emotional reactions of the music listeners may cause physiological changes as well as neurological benefits. However, any musician should avoid being influenced by an algorithmic recommendation presented in the research. The most prominent aspect for any musician is to stay true to their own musical style and to be novel, honest, and sincere. The main objective of this research work is to provide precise music predictions to the artists before they launch their music in the market. This work also aims to improve the accuracy of the prediction models using some feature-engineering-based techniques to manipulate the Spotify dataset and predict whether the song will be a hit or not. The proposed work leverages state-of-the-art machine learning models, including Linear Regression, Lasso Regression, Ridge Regression, Elastic Net, Random Forest, GBM, and neural networks to explore the predictive factors influencing the inclusion of a particular music track in the featured Spotify Hit-50 list. Furthermore, it offers descriptive statistical analysis of the features that can help in popularity prediction.