2
$\begingroup$

Let's say I have a dataframe with one dependent continuous variable and multiple independent categorical and continuous variables. I want to apply linear regression (using R language in my case). The results include that : R2 = 0.45 and R2 adjusted = 0.44.

When applying the cross validation method to improve the predictibilty of the model I created, the more I increase the number of k folds, the more the value of R2 increases. Does that indicate a mistake I am making or it is just normal to have such a result and I have nothing to worry about ?

I have a reproducible example that I can upload here, but I am not sure if I can do that here on this platform.

$\endgroup$

1 Answer 1

1
$\begingroup$

That sounds like a reasonable result. Increasing the number of folds increases the size of the training set used for each fold. A model trained on a large training set will often be more accurate than one trained on a small training set, as it is less likely to overfit. Obviously there are exceptions - a highly biased model may not improve with a larger training set.

$\endgroup$
1
  • $\begingroup$ Thank you so much $\endgroup$
    – An116
    Commented Jan 15, 2023 at 8:06

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.