Quote:
Originally Posted by John Pancoast
Well, if a model produces "right" results, by that very action it is then "right".
After all, that's the end result desired or should be. I.e., better to have that vs. a "right" model that produces "wrong" results.
But I understand what you're saying. Interesting work and effort you've done, thanks !  |
A model is right if it produces right results in
all studied cases, not if it's "only"
almost all cases.
Also, a model that produces wrong results cannot, by definition, be a right model.
But well... Thanks for your support. ;]