great video! I have a problem where I've classified a clipped raster (watershed) and now wish to generate some accuracy assessment points using the create accuracy assessment points tool. However, when I run the tool, it generates points in the NoData extent beyond the boundary of my watershed. Reclassifying or IsNulling the NoData values didn't help because all it does is create a new class in my classified raster and hence points are still generated throughout the rectangular extent of my study area. Any pointers would be much appreciated.
If I want to create a ROC curve, is it possible to only use the results from the confusion matrix? Can you please make a tutorial of making a ROC curve? I don't know what to do
great video! I have a problem where I've classified a clipped raster (watershed) and now wish to generate some accuracy assessment points using the create accuracy assessment points tool. However, when I run the tool, it generates points in the NoData extent beyond the boundary of my watershed. Reclassifying or IsNulling the NoData values didn't help because all it does is create a new class in my classified raster and hence points are still generated throughout the rectangular extent of my study area. Any pointers would be much appreciated.
select Classified as target field & set number of random points as 100 to 300. Sampling strategy can be random.
If I want to create a ROC curve, is it possible to only use the results from the confusion matrix? Can you please make a tutorial of making a ROC curve? I don't know what to do
Need to check some research papers
Is this process only for LuLC classification?
yes