I carried out a grid search on my xgboost and varied the parameters below. I noticed in my grid search results that the train score is very high, e.g. 0.99999 and my test scores are more modest around 0.72. Since there is such a big difference between my train and test score does this indicate over fitting?

enter image description here


1 Answer 1


I assume your score is a metric like accuracy, f1-score etc. If that is the case, yes, it quite certainly indicates overfitting. 28% accuracy difference between train and test/validation is very high, and it means the model didn't generalise as well for the test set as the training set.

  • $\begingroup$ what if i am also doing a grid search on say brier score, how can i tell over fitting? $\endgroup$
    – mathella
    Commented Aug 10, 2020 at 16:21
  • $\begingroup$ It still is large. It's also interesting to see that your scores don't change wrt hyperparameter changes $\endgroup$
    – gunes
    Commented Aug 10, 2020 at 19:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.