Perform the following steps to retrieve the minimum estimation errors via cross-validation with the boosting method:
- First, you can use boosting.cv to cross-validate the training dataset:
> churn.boostcv = boosting.cv(churn ~ ., v=10, data=trainset,
mfinal=5,control=rpart.control(cp=0.01))
- You can then obtain the confusion matrix from the boosting results:
> churn.boostcv$confusion Output Observed Class Predicted Class yes no no 119 1940 yes 223 33
- Finally, you can retrieve the average errors of the boosting method:
> churn.boostcv$error Output [1] 0.06565875