gbm - R - Caret - Using ROC instead of accuracy in model training -
gbm - R - Caret - Using ROC instead of accuracy in model training -
hi name abhi , using caret build gbm trees based model. instead of accuracy utilize roc metric
here code have far
mytunegrid <- expand.grid(n.trees = 500,interaction.depth = 11,shrinkage = 0.1) fitcontrol <- traincontrol(method = "repeatedcv", number = 7,repeats = 1, verboseiter = false,returnresamp = "all",classprobs = true) mymodel <- train(cover_type ~ .,data = modeldata,method = "gbm",trcontrol = fitcontrol,tunegrid = mytunegrid,metric='roc')
however when run code warning
warning message: in train.default(x, y, weights = w, ...) : metric "roc" not in result set. accuracy used instead.
how forcefulness model utilize roc instead of accuracy. doing wrong here?
here link github project source code? https://github.com/rseiter/practicalmlproject/blob/master/multiclasssummary.r
r gbm
Comments
Post a Comment