33. Feature engineering and data-centric AI

Create useful features, manage missing values and imbalance, inspect labels, and improve results by improving the dataset rather than only the model.