A Genetic Algorithm for optimization of selection algorithms in High Energy Physics

Paper: 125
Session: F (talk)
Speaker: Organtini, Giovanni, University of Rome
Keywords: algorithms, analysis, software tools


A Genetic Algorithm for optimization of selection algorithms
in High Energy Physics

Giovanni Organtini
Dip.to di Fisica Universita' di Roma III - Via della Vasca
Navale, 84 00146 Roma (Italy)

Abstract

A genetic algorithm approach is employed to write a program
that finds the best point in the multiparameter space of
the variables used for the selection of events in high energy
physics. In this paper I show the working principle of the
algorithm and how it helps not only in finding optimum values
for cuts, but also in finding the redundant variables of
the selection reducing the contribution of them to the
systematic error.