XGBoost giving an error

I desperately need your help, my XGBoost model is failing to run and showing the following error:

“ERROR XGBoost Predictor 0:208 Execute failed: The feature buffer is empty.”

How can I solve the problem?

Hi,
can you make sure that the columns coming into the predictor are the same as the ones being used by the learner? This includes column names and types. Otherwise it might be helpful to see the KNIME log file, which you can get by going to View -> Open KNIME Log… There might be more information about the problem in there.
Kind regards,
Alexander

2 Likes