Hello,
I would like to know if we can use directly the results of aicrowd for the project to evaluate the accuracy of our models?
Or should we consider cross-validation only ?
it's recommended that you do your own validation as well, this way you can find good choices much faster, not limited to 5x day, and so can potentially reach better accuracy. but as you prefer.
from a reproducibility perspective, either is fine
Using the feedback of aicrowd
Hello,
I would like to know if we can use directly the results of aicrowd for the project to evaluate the accuracy of our models?
Or should we consider cross-validation only ?
it's recommended that you do your own validation as well, this way you can find good choices much faster, not limited to 5x day, and so can potentially reach better accuracy. but as you prefer.
from a reproducibility perspective, either is fine
Add comment