1.2: Process of Training a Machine Learning Model

There are 2 steps for training a machine learning model for structured data. Feature engineering is the bare minimum needed for developing a model using a dataset. Feature selection is performed on a need basis if the features created through feature engineering can t help us develop a model of desirable predictive power.

After these 2 steps are performed, we can also perform hyperparameter tuning and ensemble learning. Although hyperparameter tuning and ensemble learning are not the book's focus, hyperparameter tuning and ensemble learning are briefly discussed in sections 11.1 and 11.2 in chapter 11.