# Is it possible to change misclassification cost in caret?

0 down vote favorite

I have an imbalanced dataset to work with (with about 20-fold positive examples than negative). I know several solutions to deal with this type of data (under/oversampling, optimizing for AUC, etc.) but I would like to try changing the misclassification costs for the two groups. However, I cannot figure out the proper way to do this in caret. Do you have any idea how to do this?

Thanks for any help!

I would have to read more carefully: train has a weights parameter that do this.