1

Different seeds give different predictions, but if you grow a sufficiently large number of trees will the predictions for different seeds converge?

Mike
  • 11
  • 1
  • 2
    Converge $\ne$ deterministic. – Tim Jan 20 '20 at 22:39
  • This question is basically answered here. https://stats.stackexchange.com/questions/348245/do-we-have-to-tune-the-number-of-trees-in-a-random-forest/348246#348246 Increasing the number of trees is effectively just increasing the number of random trials, so we can apply the usual convergence results, so we know that predictions will tend toward a mean value. – Sycorax Jan 21 '20 at 03:00

1 Answers1

0

I would guess not, because the hyperparameter ‘mtry’ still injects randomness into the features that are included during a particular run, not to mention the randomness in the cases selected during cross validation. If anything I would suspect it becomes ‘more deterministic’ as mtry approaches # of features.

HEITZ
  • 1,682
  • 7
  • 15