What is feature engineering?
Feature engineering is the process of transforming and constructing features in the dataset. It allows increasing the accuracy of the model by adding new relevant features or by combining several existing features into one.
Feature engineering is one of the most challenging steps in a machine learning workflow. It must be done with extreme care since incorrectly constructed features may result in a decreased accuracy and interpretability of the model.
Why is it important?
Feature engineering can help researchers to understand data more profoundly and to extract valuable information from it. When used correctly, feature engineering can significantly increase the model’s performance. At the same time, it allows optimizing the model training process by using only the relevant features.