#11




Re: Exercise 3.4
is not , but that is close. Recall
Quote:
__________________
Have faith in probability 
#12




Re: Exercise 3.4

#13




Re: Exercise 3.4

#14




Re: Exercise 3.4
You are right, is an matrix. But its trace is not . You may consider looking through Exercise 3.3, and in particular, part (d) should be helpful.
__________________
Have faith in probability 
#15




Re: Exercise 3.4
For part (c), would the result of (yhat  y) (from part b) be Ein(wlin) in terms of epsilon, since (yhat  y) is the insample error?

#16




Re: Exercise 3.4
y and yhat are vectors. The normsquared of (yhat  y) divided by N is the insample error.
__________________
Have faith in probability 
#17




Re: Exercise 3.4

#18




Re: Exercise 3.4
Hi James,
I am slow in responding this term as I am attending to the edX forum, but here are my quick comments: For part (a), why is (what happened to the added noise)? For part (b), your formula is correct.
__________________
Where everyone thinks alike, no one thinks very much 
Thread Tools  
Display Modes  

