Random forest old parameters

Hello everyone. I have done some classification some months ago. Now I can’t get the same results for the same images. Does anyone know where I can see which parameters did I use in my classification (number of trees and number of training samples)?? Thanks!

a reasonable answer is given here: Maximum Likelihood Classifier results are different each time when re-run on the same training data

Edit: I didn’t see that you are using a random forest classifier. @MCG is right - every result will be slightly different. But they become more similar with an increasing number of trees.

You will never have the same classification using RF (execpt if you use the same forest to be applied to the same data, with same training, etc.) beacuse randomness is part of how this algorithm works.


Thank you for your answers. I am aware that i won’t get the same results. But can I see which parameters did I use in the previous classification - number of trees and number of training samples? When I open my classification result, is there anywhere where this is written? Thanks!

yes, this information is stored in your product under metadata > processing graph.
The last node is the writing of the output file, the one before is the classification tool. You can open it and see the used parameters: