Formalism for Simulation-based Optimization of Measurement Errors in High Energy Physics

TL;DR

This paper introduces a formalism for directly estimating measurement errors in high energy physics using simulated data, eliminating the need for data fitting and background modeling, thereby streamlining event selection optimization.

Contribution

The paper presents a novel formalism for error estimation based on signal probability density functions and large simulations, improving efficiency in high energy physics data analysis.

Findings

Formalism allows direct error evaluation without data fitting.

Application demonstrated in CP violation measurement in B decays.

Implications for selecting event variables in data analysis.

Abstract

Miminizing errors of the physical parameters of interest should be the ultimate goal of any event selection optimization in high energy physics data analysis involving parameter determination. Quick and reliable error estimation is a crucial ingredient for realizing this goal. In this paper we derive a formalism for direct evaluation of measurement errors using the signal probability density function and large fully simulated signal and background samples without need for data fitting and background modelling. We illustrate the elegance of the formalism in the case of event selection optimization for CP violation measurement in B decays. The implication of this formalism on choosing event variables for data analysis is discussed.