In this project we develop a prediction system that estimates the popularity of a song. The flow of this work starts with the manipulation of the dataset: ''Spotify Track Dataset''. The training and estimation parts, together with k-fold nested cross validation and kernel functions, are developed from scratch and compared with the scikit-learn operations.
- Ridge regression
- On numerical features
- On all features
- Kernel ridge regression
- K-Fold cross validation
- On Ridge regression
- On Kernel ridge regression