How many models would we need to fit in order to exhaustively obtain theR

M

S

P

E

RMSPE

foreveryexcluded chunk of sizen

k

k

n

?

?

The method of LOOCV is not random because you t all 'n. possible models that leave out a single observation. The method of k-fold CV is random because we randomly split the data into '6: chunks. How many models would we need to t in order to exhaustively obtain the RMSPE for every excluded chunk of size E? Note: the following answers are all "combinations" (or "n choose x")