Feature engineering is essential to applied machine learning, but using domain knowledge to strengthen your predictive models can be difficult and expensive. To help fill the information gap this complete guide teaches beginning-to-intermediate data scientists how to work with this widely practiced but little discussed topic.