top of page
Predictive Analysis on Song Ratings Competition,
Hosted on Kaggle
This project put our R skills to the test. It involved performing steps of predictive analysis such as exploring and tidying data, fitting competing models, selecting features, and tuning model hyper-parameters, followed by interpreting and presenting the results.
Overview
The competition involved predicting the popularity rating of songs depending on their features from a set of 20,000 popular songs and their auditory features that were handed out to us. Using various skills and methods learned through R, my analysis resulted in a root mean squared error of 15.39157 using a linear regression model.
Take a look at the project below.
bottom of page