In our study, we want to calculate variable importance based on the model created with multiple classes. But we could not find any package or code directory that does this. Can you help with this?
caret has model-based importance measures and, when the importance code allows it, it returns importances for each class. I don't think that this is implemented in very many cases though.
This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.
If you have a query related to it or one of the replies, start a new topic and refer back with a link.
Is there a general method for obtaining model-based variable importance for a predictive model in any package in R? Not all models have variable importance in Caret. Thank you in advance for your assistance.