Performing Feature Scaling

Many machine learning algorithms are sensitive to the variable scale. For example, the coefficients of linear models depend on the scale of the feature – that is, changing the feature scale will change the coefficient’s value. In linear models, as well as in algorithms that depend on distance calculations such as clustering and principal component analysis, features with larger value ranges tend to dominate over features with smaller ranges. Therefore, having features on a similar scale allows us to compare feature importance and may help algorithms converge faster, improving performance and training times.

Scaling techniques, in general, divide the variables by some constant; therefore, it is important to highlight that the shape of the variable distribution does not change when we rescale the variables. If you want to change the distribution shape, check out Chapter 3, Transforming Numerical Variables.

In this chapter, we will describe...