L1 and L2 regularization

Another way to prevent overfitting is to modify the loss in order to include regularization as one of the training objectives. The L1 and L2 regularizers are prime examples of this.