This repository contains code and resources about a study focused on developing a linear regression model using R programming to predict song energy levels based on various audio features. The dataset comprised songs from the Spotify Top 100 playlist of 2018, with predictors including danceability, valence, time signature, and other relevant audio features.
- Utilization of a Spotify Top 100 2018 dataset to predict song energy levels
- Examination of audio features such as loudness, danceability, and acousticness as predictors of energy
- Identification of significant predictors indicating higher energy levels in specific audio features
- Evaluation of the model's performance, showcasing a good fit with a significant R-squared value, explaining a substantial portion of energy variation based on selected features
- Insights beneficial for music recommendation systems and playlist curation
- Enables personalized song recommendations based on desired energy levels
- Facilitates the creation of playlists catering to specific energy preferences