I am running Random Forest Classifier using ArcGIS 10.4. How do you assess the error of the tree ensemble, which is typically done using out of bag (OOB) estimate? I opened the .ecd file and I see "CrossValidateRate" and a value. What is this rate referring to and is it the equivalent of OOB estimate? Are bootstrap samples used to derive "CrossValidateRate"?
According to Breiman:
"In random forests, there is no need for cross-validation or a separate test set to get an unbiased estimate of the test set error. It is estimated internally, during the run, as follows:
Each tree is constructed using a different bootstrap sample from the original data. About one-third of the cases are left out of the bootstrap sample and not used in the construction of the kth tree".