Data Normalization
From Open Risk Manual
Definition
Data Normalization is a loose term that refers to data transformation activities, typically in preparation for the development of a Risk Model that aim to improve the informational content of the data by removing manifestly extraneous influences.
In a broad sense it may denote:
- the reduction of datasets to relatively uniform or homogeneous pools
- reduction of unrelated Data Outliers
- addressing Seasonality Effects
In a more narrow sense it may mean the scaling of data to conform to a normalized scale.