0
$\begingroup$

You’ve built a random forest model with 10000 trees. You got delighted after getting training error as 0.00. But, the validation error is 34.23. What is going on?

$\endgroup$
0

1 Answer 1

1
$\begingroup$

That means you're overfitting your model such that it gives no training error, but unable to generalize correctly to OOB.

One possibility is that you have very complicated trees in your model (e.g. no pruning). Another possibility is that you have a "cheating" feature in your data set (and that's also your only feature), for example, Case ID.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.