It is the process of reducing random variables under consideration. Dimensionality reduction can be classified as feature selection and feature extraction.
Feature selection tries to find the subset of input variables, while feature extraction begins from an initial set of measured data and builds derived values.
