ST4248 Project
1 Introduction
Music Popularity Prediction Using Machine Learning Techniques
1.1 Project Description
In this project, we aim to predict the popularity of a song based on its features using various machine learning models, such as Random Forest, Linear Regression, Support Vector Machines (SVM), and XGBoost. We will preprocess the data, perform Principal Component Analysis (PCA) for dimensionality reduction, and apply hyperparameter tuning to find the optimal parameters for each model. The performance of each model will be evaluated using metrics such as Mean Squared Error (MSE), Root Mean Squared Error (RMSE), R-squared, and accuracy. By comparing the results of the different models, we will determine the most effective approach for predicting a song’s popularity. This analysis can be used to better understand the factors that contribute to a song’s popularity and could potentially be employed by music producers, record labels, or streaming platforms to predict the success of new releases.
1.3 Motivation
In the highly competitive music industry, understanding what makes a song or artist valuable is crucial for marketing strategies, album releases, and aligning efforts with public interest. Spotify, one of the largest streaming services, reaches over 191 million active users and publishes several charts, including a daily “Top 50” ranked by the number of streams. For artists, knowing when their work will become popular on the platform is essential for boosting marketing efforts and steering efforts towards more promising cases. Therefore, predicting a song’s popularity using musical characteristics can be of great interest to artists and the music industry.