When I optimise the XGBoost model hyper-parameters, the first observation is OK, but from the 2nd observation onwards I simply get the same value being returned over and over.
I have tried different optimisation functions, e.g., AUC, f1-score, accuracy, etc. Always the same problem.
I have also tried all the solutions from the linked Issue above, where possible.
Hi @yanyachen,
I have encountered the same bug in the R implementation as in the python version: https://github.com/fmfn/BayesianOptimization/issues/10 .
When I optimise the XGBoost model hyper-parameters, the first observation is OK, but from the 2nd observation onwards I simply get the same value being returned over and over.
I have tried different optimisation functions, e.g., AUC, f1-score, accuracy, etc. Always the same problem.
I have also tried all the solutions from the linked Issue above, where possible.
Please advise and thanks for your time.