Principal component analysis (PCA)

Technique used to find the most valuable parts of all of the variables in a dataset and to then transform the original variables into a smaller set of linear combinations.