The gbm3 library contains a function to calculate predictions from cross validated generalized boosting models. If I understood that correctly, predictions should be made for each cv fold. However, it is unclear how a GBMCVFit object containing CV gbm models is made or stored? Which functions produces as GBMCVFit object to be used for prediction?
The gbm3 library contains a function to calculate predictions from cross validated generalized boosting models. If I understood that correctly, predictions should be made for each cv fold. However, it is unclear how a GBMCVFit object containing CV gbm models is made or stored? Which functions produces as GBMCVFit object to be used for prediction?