Feature Engineering

---

Feature engineering is so important to how your modeling pipeline performs, that even a simple pipeline with great features can outperform a complicated one with poor ones. In fact, feature engineering has been described as easily the most important factor in determining the success or failure of your predictive pipeline. Feature engineering, sometimes, really boils down to the human element in machine learning. How much you understand the data, with your human intuition, creativity, and domain expertise, can make the difference.

When conducting an end to end Machine Learning project, after exploring and preprocessing the data it is essential to think of feature engineering. It consists of creating new feature(s) based on the features that already exist in the dataset (or that can be merged in from an external/third-party dataset). These features can then be used in during training the pipeline.