No. 14 (00084)

Family name : Jankowski
Given name : Ernest
Affiliation : University of Alberta
Abbreviation : 
E-mail address : ejankows@phys.ualberta.ca
Title : Optimal Jet Finder
Authors : D.Yu. Grigoriev, E. Jankowski, F.V. Tkachov
Abstract : 
The Optimal Jet Finder (OJF) is a program that identifies jets in
hadronic final states of particles collisions. It is an implementation
of the Optimal Jet Definition.

The final jet configuration is determined by the global energy flow in
the event , which distinguishes this jet definition from cone or binary
recombination algorithms such as kT. kT merges only two particles at a
time, whereas OJF finds the final jets by minimizing some global
function of the momenta of all input particles. The idea of using the
global structure of an event for defining jets is not new, but it
involves a non-trivial optimization problem. To the best of our
knowledge, OJF is the first program of this kind working sufficiently
fast to be of practical use in data analysis.

The jet definition avoids the ambiguities of cone algorithms involving
seeds due to emission of soft particles or collinear spiting of
particles. It handles overlapping jets without any additional arbitrary
prescription.

OJF is significantly faster in comparison with kT (or any other binary
algorithm) if a large number of input particles/calorimeter cells has to
be processed. The average time to analyze an event depends linearly on
the number of input cells whereas for kT this dependence is cubic. This
feature possibly makes OJF appropriate for studies at the detector level
without the preclustering step which introduces further uncertainties,
especially that it is difficult to incorporate the preclustering into
theoretical calculations.

Besides the jet configuration, Optimal Jet Finder also evaluates
additional numerical characteristics of the jet configuration found
(dynamical width of jets and soft energy). Those characteristics can
facilitate the construction of (quasi-) optimal observables in
statistical problems. This is a further unique property of OJF in
confrontation with kT or cone algorithms.

I will describe the essential features of the Optimal Jet Definition. I
will explain how it arises from the application of optimal observables
to the analysis of hadronic final states via a jet algorithm. I will
give the important details of the implementation. I will bring out the
test applications done so far and prospects for its use in high energy
data analysis.

Related publications:

(1) Fyodor V. Tkachov,
    A Theory of Jet Definition,
    Int.J.Mod.Phys.A17:2783-2884,2002; hep-ph/9901444

(2) D.Yu. Grigoriev, E. Jankowski, F.V. Tkachov,
    Towards a Standard Jet Definition,
    Phys. Rev. Lett. in print; hep-ph/0301185

(3) D.Yu. Grigoriev, E. Jankowski, F.V. Tkachov,
    Optimal Jet Finder,
    Comp. Phys. Commun. in print; hep-ph/0301226

(4) D.Yu. Grigoriev, E. Jankowski, F.V. Tkachov,
    A Monte Carlo Test of the Optimal Jet Definition,
    Lake Louise 2003 proceedings; hep-ph/0306224