“So the best hyperparameters give a model where the features are scaled but not centered and the final model is regularized.”
what is the meaning of not centered and the final model is regularized. Did you come to these conclusions from these results ?
{‘columntransformer__num_preprocessor__with_mean’: False,
‘columntransformer__num_preprocessor__with_std’: False,
‘logisticregression__C’: 0.17169565852473864}
Centered relates to with_mean=True and scale related to with_std=True.
However, the conclusion is not in line with the results that you get but I think that we already have a post where we acknowledge this problem.
Here, your results show that you have a model that does not centre nor scale.