Week 7 — Lesson 5: Regularization and early stopping

3 min

In Week 6 a depth-20 tree scored 1.000 on the fit months and 0.726 on validation. The NorthPeak neural network, with 1,025 weights for 10,720 rows, can do the same: give it enough epochs and it fits the noise of the training rows. Two tools hold it back. Regularization penalises large weights through alpha; early stopping watches a held-out score and stops when it stalls. This lesson trains the same (32, 16) network with and without these tools and compares the test AUC, 0.945 against 0.923.

Preview — the rest of the lesson is for enrolled readers.

Already enrolled with a code?

Your access is tied to your account, not to this link. Sign in with the same email you used in class: your course is waiting, no need to enter the code again.

This lesson is part of the “Week 7 — Clustering and neural networks” module

The first modules of the course are open to everyone. For the rest you have three options: buy this course once and for all, subscribe, or enter the code handed out in class.

Back to the outline