3 ms·
Because the model doesn't overfit and that is good. You don't want your model to mimic your trainning data. Otherwise, it would performe poorly with new data.
by manceraio 6y ago
Because the model doesn't overfit and that is good.
You don't want your model to mimic your trainning data. Otherwise, it would performe poorly with new data.