Dear Andrzej,

Just want to check with you if it is possible to reopen the submission portal based on the full testdata sets for some time until the deadline of paper submission so that so we can explain more with the models and feature selecting in the paper, as you know most time we data scientists just work on "try and see" strategy and relies on luck :-).

And as for me, I did one distance shifting mechanisms instead of the simple average when do merging models, it really works in the level 1 datasets and can improve the score a lot, while since xgboost dominates the results, then it is not really promising as before, so I also would like to investigate how this mechanisms work on the whole test datasets.

However if you think it will increase your team a lot workload, then just ignore this request then.

Thanks again for you and your team's hard work and efforts.

Warmest regards

David