I've set a breakpoint after an apply model node in a decision tree model with cross validation. After executing the breakpoint correctly, the model won't proceed. It just keeps repeating the breakpoint. I don't have this problem with models that have no subprocesses. The XML follows. Any ideas?
The cross validation operator will execute the subprocess 10 times (or whatever the 'number of validations' parameter is set to).
Each iteration trains a model on a subset of the data and tests it on another subset. The model is discarded but the performance is kept for later. The iterations repeat this for different partitions of the data.
At the end, the performances are averaged. The model produced by the operator is built on all the data and the average performance can be regarded as an estimate of how this model would perform on unseen data.