Using boosted decision trees for tau identification in the ATLAS experiment

The ATLAS detector will begin taking data from $p$-$p$ collisions in 2009. This experiment will allow for many different physics measurements and searches. The production of tau leptons at the LHC is a key signature of the decay of both the standard model Higgs (via H $\rightarrow \tau \tau$) and SUSY particles. Taus have a short lifetime ($c \tau= 87$ $\mu$m) and decay hadronically ~65\% of the time. Many QCD interactions produce similar hadronic showers and have cross-sections about 1 billion times larger than tau production. Multivariate techniques are therefore often used to distinguish taus from this background. Boosted Decision Trees (BDTs) are a machine-learning technique for developing cut-based discriminants which can significantly aid in extracting small signal samples from overwhelming backgrounds. In this study, BDTs are used for tau identification for the ATLAS experiment. They are a fast, flexible alternative to existing discriminants with comparable or better performance.