I have a simple classification problem with 33 possible predictors (all categorical variables). I wanted to determine what the best predictors are to include them in a Naive