# Standardization in Machine Learning

Normalization usually helps accelarate convergence of algorithm.

## 0-1 standardization (MinMaxScaler)

MinMaxScaler standardization is one of the most widely recognized approaches to standardize information. For the variables, the base estimation of that element gets changed into 0, the most extreme worth gets changed into a 1, and each other worth gets changed into a decimal somewhere in the range of 0 and 1. For the responses, the elements can be standardized to have unit \(\|\cdot\|_2\) norm. In many applications, the responses \(y\) also retain the original scales.

## Z-score standardization (StandardScaler)

The StandardScaler is used for standardizing scores on the same scale where the variables are standardized to have unit \(\|\cdot\|_2\) norm.