3 ms·
Possibly silly question: what would happen if one were to extrapolate? I mean, using the trained model in a larger grid. I should do it myslef but have no back
by pfortuny 8y ago
Possibly silly question: what would happen if one were to extrapolate? I mean, using the trained model in a larger grid.
I should do it myslef but have no background in applications.
- mariofilho 8y agoIt will very likely fail to predict new data. The model can approximate the interactions over the range it saw on training, but we need to show it larger ranges if we want to predict for a bigger grid.
- hopler 8y agoSo, Betteridge's Law of Headlines is supported.