In this time, I work on the pleiades’ data and I want to make a supervised classification with random forest. My result is not bad but there are a lot of pixels with NaN values.
I have six classes in my vectors data and they are all represents in my classification.
Maybe it’s an other classe which is not represented ?
Or the software can’t take a decision because two variables are too close ?
Thanks for your help !
Have a good day.
this can have several reasons:
- pixels are classified as NaN because their confidence is too low
- pixels are classified as NaN because one of the input bands is NaN at this location
- the training areas for this class are not large enough (or contain too inhomogenous signatures
Have you seen these discussions?
Ok It’s that I thought…
I saw the first one !
Thanks for the help !