Homepage › Solution manuals › Yaser Abu-Mostafa › Learning from Data › Exercise 4.8

Exercise 4.8

Answers

For each model $H_{m}$ , $g_{m}^{−}$ is independently learned of from the validation set. When we take the expectation w.r.t. validation data set, the $g_{m}^{−}$ doesn’t change, the expectation depends only on $x_{n}$ , as in the derivation of equation (4.8).

So I think $E_{m}$ is an unbiased estimate for the out-of-sample error $E_{out} (g_{m}^{−})$ .

niuers

2021-12-08 09:30

Exercise 4.8

Answers

Comments

Add answer