Part 02 · Machine Learning

Overfitting and Regularization

Memorising the training set is not learning. The bias-variance trade-off, reading learning curves, and L1, L2 and early stopping to keep a model honest.

⏱ 1-2 hours · Station 13 of 92