Polynomial regression model

• Adding powers to the linear regression model in order fit non-linear data
  • Also adds all combination of features up to the given degree, giving the ability to capture interactions
• The more powers we add to the polynomial model the better we will be able to capture the data, but that will lead to overfitting
  • One way is to use cross-validation metrics to ensure we are not overfitting
  • Another way is to plot [[learning curves]] of the model