Comparison of Cut-Based and Matrix Element Method Results for Beyond Standard Model Quarks

TL;DR

This paper compares cut-based and matrix element methods for estimating the mass of a hypothetical fourth-generation quark in proton-proton collisions, demonstrating the superior accuracy of the matrix element approach especially with limited data and higher signal-to-background ratios.

Contribution

It introduces a comparative analysis of two mass extraction techniques applied to a beyond Standard Model quark, highlighting the advantages of the matrix element method in specific experimental conditions.

Findings

Matrix element method yields more accurate mass estimates with fewer events.

The advantage of the matrix element method increases with higher signal-to-background ratios.

Even with limited data, the matrix element method outperforms the cut-based approach.

Abstract

In this work, two different methods for extracting the mass of a new quark from the (pseudo) data are compared: the classical cut-based method and the matrix element method. As a concrete example a fourth family up type quark is searched in p-p collisions of 7 TeV center of mass energy. We have shown that even with very small number of events, Matrix Element Method gives better estimations for the mass value and its error, especially for event samples in which Signal to Background ratio is greater than 0.2.