If I use Weight of Evidence to transform categorical variables, do I still need to inform their indexes to Catboost

Question

I'm using Weight of Evidence (WOE) to encode my categorical features. Do I still need to inform Catboost that they are categorical features by using cat_features parameter?

Answer 1

The goal to encoding is to transform categorical into numerical so that an algo can learn on them. So the general answer would be no, after encoding into numerical you shouldn't declare them as categorical anymore.

However, as its name suggest it, catboost is designed to handle categorical... so for catboost you probably want to use categorical. In this case you shouldn't need to encode categorical at all... There are some very specific exceptions (over-fitting), and even in that case you might want to start with other parameters - min data in leaf for exemple before getting into relatively complex encoding. WoE encoding makes sense for simpler models (basically it should work well with logistic regression by design).