|
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
|