Generalization

Reading tips: lecture notes

Can the network generelize to examples that have never been seen?

What is meant by overfitting?

How can overfitting be avoided?

What is meant by cross validation?