1
$\begingroup$

Is it correct that each feature is only used once in one decision tree ? Then, if a feature (the most decisive for the classification) determines a classification for a range of values, can the algorithm detect that in any way or will it fail ?

To be more clear, let's say the classification is to classify people as teenager or not. Then the determinant feature will be the age, and classification will be if this feature is between 12 and 18. So we don't have only one threshold (age<12) or (age>18), but two.

Can the Random Forest algorithm succeed in this case ?

Thanks for any suggestion

Improve this question
$\endgroup$

1 Answer 1

Reset to default
2
$\begingroup$

Is it correct that each feature is only used once in one decision tree?

No, if the variable is a continuous value like age, it can use threshold splits at each level. So, same feature can be used multiple times in any given branch of the decision tree.

And since the random forest is nothing but collection of decision tree it follows the same principle.

Reference

Improve this answer
$\endgroup$
1
  • $\begingroup$ Ok, thanks a lot for the answer ! $\endgroup$
    – ZiGaelle
    Commented Nov 25, 2020 at 14:25

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.