Redundant Variables and Machine Learning

In our previous lesson, we observed (in the day column) that if three days are 0, the fourth must be 1. We don’t need an extra column in all the cases; these are redundant. Therefore, it’s recommended to drop the redundant variables first. Otherwise, lasso reduces them to zero, even with a mild regularization strength.

Drop redundant variables

Let’s drop all the redundant variables and look at the heatmap again.