Types of feature scaling

Standardization:

Standardization is the simplest form of scaling, in which all the values are standardized to have a mean of zero and a standard deviation of one. For example, if you had a dataset with two variables (age and height), then you would calculate their means and standard deviations before performing any statistical tests on them.

Normalization: 

Normalization method also involves calculating and standardizing every value in your dataset before performing any statistical test but instead finds its median value as well as its mean value. It then determines whether those two values differ significantly from each other based on how far apart they fall from that first reference point; if so then it assumes that there is something wrong with either one or both of them so as not to draw conclusions about what’s happening within our sample population (e.g., “My kid might be taller than average because he grew faster than most kids his age”).

Feature scaling is a technique to improve the accuracy of machine learning models. This can be done by removing unreliable data points from the training set so that the model can learn useful information about relevant features. Feature scaling is widely used in many fields, including business analytics and clinical data science.