I have run Random forest regression with python and I have the fear that I haven't done it correct.
I have original image that has 3 different bands (each pixels has 3 values) and I want to try to see if I can predict the first value using the two others.
For that I used Random Forest regression, I have created train and test and fit the model:
then after I checked the prediction on the test and saw it was good, I wanted to use the same model in order to predict all the values of all the pixels in the image, so I did that:
*data includes all the pixels of the image.
My question : is this the right way to do this? can I just predict all the pixel values just by using the rf.predict after fit it with the train and test sets? Or am I missing here some step that should be taken?