is it allowed to change or add features to train and test dataset and run model on the reconstructed one?
That shouldn't be a problem. It's a valid feature engineering step. Secondly, the hosts should be interested in the submission format and working code, as per my opinion.
thank you @Mandar_Thosar
Now I got it because afterall its what predictions what matters the most.
Obviously when doing feature engineering we have to keep train and test data of same dimension
That shouldn't be a problem. It's a valid feature engineering step. Secondly, the hosts should be interested in the submission format and working code, as per my opinion.
thank you @Mandar_Thosar
Now I got it because afterall its what predictions what matters the most.
Obviously when doing feature engineering we have to keep train and test data of same dimension