Closed dluts closed 1 month ago
We have two categorical features that are not booleans, maybe we should apply a onehotencoder.
This has been added to the consolidated notebook.
Okay, by doing this discovered that with the imputation it was not only the imputer that messed with the categories, but also the scaler.
Limited application to train/test for linear models and SVM
We have two categorical features that are not booleans, maybe we should apply a onehotencoder.
This has been added to the consolidated notebook.