Definition
Techniques used to reduce the number of features or variables in a dataset while striving to preserve the most important information and underlying structure.
In short: Techniques used to reduce the number of features or variables in a dataset while striving to preserve the most important information and underlying structure.