Min-max scaling

Min-max scaling, also known as normalization, scales the values of a variable to a specific range, typically between 0 and 1. Min-max scaling is useful when you want to ensure that all values in a variable fall within a standardized range, making them directly comparable. It is commonly employed when the distribution of the variable is not assumed to be normal.

Let’s have a look at the formula for calculating min-max scaling:

X _ scaled =(X − X _ min) / (X _ max − X _ min)

As you can see from the formula, min-max scaling preserves the relative ordering of values but compresses them into a specific range. One thing to note here is that it is not a way to deal with outliers and if outliers exist in the data, these extreme values can disproportionately influence the scaling. So, it is a good practice to deal with outliers first and then proceed to the scaling of features.

Scaling to a specific range is a suitable approach when the following...