Improvement of signal significance of Ds± meson's signalvia using artificial neural network

The aim of the analysis is to improve the significance of the Ds ± meson signal in ALEPH data which is recorded at LEP in the Ds± → π± + φ where φ mesons are reconstructed from φ →K+ K- decay channel. The analysis covers two phases. In first phase, reconstruction of particles in the decay channel is performed and input variables are stored to be used in classification. This phase is the machine learning part of the analysis. The second phase is the selection of an optimum algorithm for classification of signal and background. Improvement of significance of Ds meson's signal is performed after selection of the optimum algorithm and the evaluation of coefficients for relevant input variables which are stored in first phase. Classification is performed with the Toolkit for Multivariate Analysis (TMVA) which operates under the ROOT data analysis framework. TMVA includes many classification algorithms; the Boosted Decision Tree method is chosen as the optimum classification method.