Cascade Training Technique for Particle Identification

Abstract

The cascade training technique which was developed during our work on the MiniBooNE particle identification has been found to be a very efficient way to improve the selection performance, especially when very low background contamination levels are desired. The detailed description of this technique is presented here based on the MiniBooNE detector Monte Carlo simulations, using both artifical neural networks and boosted decision trees as examples.

12 pages and 4 EPS figures

References in corpus (15)

• Evidence for Neutrino Oscillations from the Observation of Electron Anti-neutrinos in a Muon Anti-Neutrino Beam
• Boosted Decision Trees as an Alternative to Artificial Neural Networks for Particle Identification
• Neural Network Parametrization of Deep-Inelastic Structure Functions
• Evidence for production of single top quarks and first direct measurement of |Vtb|
• Neural Network based Electron Identification in the ZEUS Calorimeter
• Search for Single Top Quark Production at DZero Using Neural Networks
• Application of neural network for photon-hadron discrimination in a preshower detector in high energy heavy ion collisions
• Neural network parametrization of the lepton energy spectrum in semileptonic B meson decays
• Separating True V0's from Combinatoric Background with a Neural Network
• Applying Rule Ensembles to the Search for Super-Symmetry at the Large Hadron Collider
• Role of Neural Networks in the Search of the Higgs Boson at LHC
• An Orientation Selective Neural Network and its Application to Cosmic Muon Identification
• Signal Confidence Limits from a Neural Network Data Analysis
• Separation of a single photon and products of the ${\pi }^0,\eta ,K_s^0$ meson neutral decay channels in the CMS electromagnetic calorimeter using neural network
• Event Selection Using an Extended Fisher Discriminant Method

Cited by in corpus (1)

• A multivariate approach to heavy flavour tagging with cascade training