1
$\begingroup$

How do we decide the learning rate in gradient boost? I can think of cross validation as one of the methods, but are there any other methods? Also, my teacher said in the class that 0.1 is a good measure for the learning rate (but did not back it with any sources).

$\endgroup$

1 Answer 1

1
$\begingroup$

It's a hyperparameter, so you pick it by tuning it, usually using cross-validation. There is no one-size-fits-all "good" value of the parameter, as it will be problem-specific. Algorithms like gradient boosting in many cases will let you get away with good-ish results even if you stick to some "default" settings, but by no means they are the optimal ones.

$\endgroup$

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.