Regularization and Overfitting

[@MadhuWall]

There is a mistake at 2:58. The error in the model for training should actually be less than the test set because it overfits to the training set and minimizes the error in the test set.