One very important step in dataset preprocessing is normalization and feature scaling. Data normalization allows our optimization techniques, specially the iterative ones, to converge better, and makes the data more manageable.
