Python Machine Learning By Example - Second Edition

Best practice 13 – documenting how each feature is generated

We have covered the rules of feature engineering with domain knowledge and in general, there is one more thing worth noting: documenting how each feature is generated. It sounds trivial, but oftentimes we just forget about how a feature is obtained or created. We usually need to go back to this stage after some failed trials in the model training stage and attempt to create more features with the hope of improving performance. We have to be clear on what and how features are generated, in order to remove those that do not quite work out, and to add new ones that have more potential.