Overfitting, Underfitting, and Model Capacity | Lecture 4

Overfitting refers to a model that models the training data too well.

Underfitting refers to a model that can neither model the training data nor generalize to new data.

Model capacity is the amount of elements that a model can fit.