Hyperparameter tuning

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view

Hyperparameter tuning

I am using xgboost hyperparameter tuning for the value of rmse. I used the
following method which returns two values ; mean which is like - 1200.12
and Best value like - 960. I guess the first value is average value and the
second is the best rmse value returned by the algorithm.

Now if we have to use Accuracy as estimation which, we know is the higher
it is, the better it will be. So do we have to use the same function but
just remove the negative sign?

For rmse ; return (- xgboost$results$rmse)

For AUC; return (xgboost$results$auc)

        [[alternative HTML version deleted]]

[hidden email] mailing list -- To UNSUBSCRIBE and more, see
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.