Newer version has a worse result


#1

I trained model using python xg offline and using biz.k11i:xgboost-predictor:0.2.1 to load model file online to predict.

result like below:

Is it version-mismatch making this happen?
Or any other reasons?

PS: In python 0.62 , all model I trained, a good offline auc will lead to a good online ctr


#2

It’s hard to tell why, since you are getting good AUC on validation dataset. Can you raise regularization parameters to improve generalization?


#3

Also look at the model dumps. Sometimes looking at actual splits can give useful insight