1
$\begingroup$

I have a set of classified data. In order to test the precision of several algorithms- I split the data into train and test sets. For the test set I choose at random 30% of the data and the rest is for the train set. I do $N$ iterations in which I split the data this way and take the average precision of the $N$ iterations.

Does this validation procedure has a specific name?

Improve this question
$\endgroup$
4
  • $\begingroup$ CrossValidation? :) $\endgroup$
    – CloseToC
    Commented Jan 31, 2015 at 10:36
  • 1
    $\begingroup$ that looks more like repeated hold out with a 70/30 split, if you are sampling without replacement, bagging if you are sampling with replacement. $\endgroup$
    – image_doctor
    Commented Jan 31, 2015 at 10:43
  • $\begingroup$ @CloseToC I need to distinguish this validation from the k-fold cross validation, so the name "CrossValidation" is not what I am searching for. $\endgroup$
    – Anton
    Commented Jan 31, 2015 at 12:51
  • $\begingroup$ @image_doctor "Repeated holdout" -> this is what I was looking for. You can add the comment as an answer. $\endgroup$
    – Anton
    Commented Jan 31, 2015 at 13:17

1 Answer 1

Reset to default
1
$\begingroup$

That looks like repeated hold trials out with a 70/30 split.

Improve this answer
$\endgroup$

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.