What to make a hybrid between GWR and XGBoost for prediction?
The GWR (Geographically Weighted Regression) is based on the OLS in formulating a prediction which considers the neighboring observations.
How to replace the OLS by the XGBoost when working with GWR?