Exploration of Jet Substructure with Energy Flow Polynomials

The Energy Flow Polynomials (EFPs) are a set of jet substructure observables which forms an overcomplete basis for all infrared- and collinear-safe jet observables. In this study, an in-depth look into the EFPs is conducted. Comparisons of EFPs distributions are made between experimental and simulated dataset; strong correlations between different EFPs are found; Principal Component Analysis is done to 53 different EFPs, where they can be reduced to 6 principal components without much loss of information; Linear Regression is done to the same set of EFPs, and also to the 6 reduced components, to find the coefficients for them to linearly combine into some traditional jet shapes. Similar performances are found between the 53 EFPs and the 6 principal components.