285443 (3) [Avatar] Offline
Wouldn't it be better to evaluate it with something like R^2 loss?
Jeff Smith (14) [Avatar] Offline
Fair point. I think that this example is a bit muddled, so I'll need to take a crack at clarifying it.

Most of the book focuses on classification not regression problems, so I don't really want to get into R^2. I think that I can probably just refactor this example to be a classification problem.