Quote:
Originally Posted by Michael Reach
I am having some trouble understanding this question, and making sure it isn't a trick. I note that in lecture 8, the professor gave an equivalent of VC dimension for linear regression  but in quotes, saying that the regression example doesn't really have a VC dimension like classifications do. Did I misunderstand? Is this question asking for the VC dimension or the "VC dimension" equivalent? Is the right answer something like d+1, or should I be choosing 0 or infinity or something?

Questions 2 and 3 are about linear classification, so the VC dimension in this case is genuine. Thank you for bringing this up, as some people may be confused after seeing Question 1 about linear regression and not noticing "decision boundary" in Question 2.