You're working on a supervised learning problem and need to decide which input features to create, transform, or exclude before training a model.
How would you discuss feature engineering choices for a predictive model?
