classification model And ClassificationPartitionedModel
4 views (last 30 days)
According to what I have learnt, I could use "CVMdl = crossval(Mdl)" to create a cross-validated classification model from a classification model. But, I'm not sure about the function of the "CVMdl".
Question: Is the "CVMdl" just for cross validation? If the result of the "kfoldLoss" is good, could I use the "Mdl"？
Vishwas Kumar on 9 Jan 2018
CVMdl is the cross validated model, you can find things like cross-validated classification losses, cross-validated classification edges, class labels predicted by the cross-validated ECOC model, etc.
And yes, based on the results of Kfoldloss, you can use Mdl further.