| Description |
A nucleus-nucleus or proton-nucleus collision at the LHC produces a tiny lump of fluid which expands into the vacuum. This is revealed by striking correlation patterns seen in data, which are understood as imprints of the fluctuations of the little fluid. Principal component analysis is a statistical procedure that extracts all of the information contained in correlations in a systematic way and expresses it in terms of fluctuations. I explain how the method works and I show recent application to LHC data by the CMS collaboration.
|