Normalization is the process of transforming numerical values into a common scale, such as 0 to 1, -1 to 1, or standardized z-scores. It helps models train more reliably by preventing features with larger numeric ranges from dominating learning.
Broadly speaking, the process of converting a variable's actual range of values into a standard range of values, such as: - -1 to +1 - 0 to 1 - Z-scores (roughly, -3 to +3) For example, suppose the actual range of values of a certain feature is 800 to 2,400. As part of feature engineering, you could normalize the actual values down to a standard range, such as -1 to +1. Normalization is a common task in feature engineering. Models usually train faster (and produce better predictions) when every numerical feature in the feature vector has roughly the same range. See also Z-score normalization. See Numerical Data: Normalization in Machine Learning Crash Course for more information.