Since there is randomness involved in the computation of a random forest classifier, it is necessary to define a random seed to get reproducible results. Is this possible to do in the current implementation of the algorithm (4.1.2 or newer)?
I don’t know if the code supports it now but I found this oldish issue and another question/answer.
please feel encouraged to edit OpenCV source to your liking and submit a pull request.
since this is using the global random generator, you can just seed that:
theRNG().state = 987654321;
I’ve attempted to include that line, as well as setting rng.state = 987654321, however I’m still getting varying results on different runs of training. Do you have any suggestions on what parts of the code I should reset the state?