I want to perform a nested cross validation for XGBoost classification model

I am struggling since last couple of days for building a workflow. I have a dataset where 1500 small molecules are there. I have calculated the morgan fingerprint. Upon that dataset I want to perform the nested cross validation for selecting the best hyperparameters. In the outer loop 10 fold cross validation will be performed and in the inner loop hyperparameter optimization followed by 5 fold cv will be done. Then atlast workflow would give the predicted performance of the best model on the whole training and test set. But I have a request, I would not use parameter optimization loop. Because I have selected hyperparameters. I want to test them only that for each loop. It would be very helpful if someone help me for that.

Hi @palsourav30,

supposing you are using Schrödinger extensions, aren’t you? If correct, maybe your post, albeit sparking my curiosity, is highly specific and better located in this category:

Apply the tag for Schrödinger to raise visibility too.