Background estimation with the ABCD method: Glossary

Key Points

Introduction
  • Data-driven background estimates are a must in situations where you cannot get a reliable estimate from simulation

  • They are also useful to validate predictions from simulations

  • The ABCD method is a common background estimated concept, based on four different regions in phase space

  • In this method, background shape in the signal region is estimated using a control region

  • Differences between the control region and signal region are accounted for by event weights called transfer factors

Control regions
  • Signal-depleted control regions are often used in background estimations

  • In the ABCD method, the background shape in signal region D is estimated using a control region C

  • To obtain estimate for a single process, e.g. QCD multijets, one can simply subtract other known (simulated) processes from data

Transfer factors
  • Transfer factors are event weights that account for the differences between a control region and the signal region

  • In the ABCD method, they are derived by comparing regions A and B, and applied to transfer the background estimate from region C into the signal region D

Summary & discussion
  • The basic concept of ABCD method is rather simple, but in practice things can get complicated

  • Careful validation of your background estimates is a necessity

  • There are several more advanced techniques in use, based on the basic concept of the ABCD method

Glossary

FIXME