Blind background prediction using a bifurcated analysis scheme

TL;DR

This paper introduces a background prediction method that uses data with a bifurcated analysis scheme, applicable to multiple sources and correlated cuts, demonstrated on a toy model and a rare kaon decay experiment.

Contribution

The paper presents a novel bifurcated analysis technique for background prediction that maintains a closed signal region and handles multiple background sources.

Findings

Effective background prediction in toy model and real data

Applicable to multiple background sources with correlated cuts

Maintains a closed signal box during analysis

Abstract

A technique for background prediction using data, but maintaining a closed signal box is described. The result is extended to two background sources. Conditions on the applicability under correlated cuts are described. This technique is applied to both a toy model and an analysis of data from a rare neutral kaon decay experiment.