clarification: ensemble seems to use only subset of methods described in README

[espinielli]

the code in csv_ensemble.ipynb states in the Loading data section

models = ['catboost', 'xgboost', 'lightgbm']#, 'mlp', 'saint']

The README states that all the model have been combined but from the line above it seems only some of them have been used for the last submission, is that the case?

[Gabrui]

Yes, the mlp and saint models had a much higher error than the other methods, and when included, their learned weights were very close to zero (<1e-10), so their influence was negligible.