Regularization (L1/L2)

Regularization adds a penalty for complexity to the loss, so the model is rewarded not just for fitting the training data, but for keeping its weights small.

This lesson is part of BitBootcamp Pro. Unlock it with Pro, or browse the free courses.