Feature engineering in ML consists of four main steps: Feature Creation, Transformations, Feature Extraction, and Feature Selection . Feature engineering consists of creation, transformation, extraction, and selection of features, also known as variables, that are most conducive to creating an accurate ML algorithm.