An improved full reconstruction tool utilizing NeuroBayes Physics - PowerPoint PPT Presentation

An improved full reconstruction tool utilizing NeuroBayes Physics In Collision 2010, Poster Session M. Feindt, M. Kreps, T. Kuhr, F. Keller, S. Neubauer, D. Zander, A. Zupanc | September 3rd, 2010 I NSTITUT F ¨ UR EXPERIMENTELLE K ERNPHYSIK , K ARLSRUHE I NSTITUTE OF T ECHNOLOGY (KIT) www.kit.edu KIT – University of the State of Baden-Wuerttemberg and National Laboratory of the Helmholtz Association

The Principle of the Full Reconstruction B factories : Center of mass energy of √ s = 10 . 58 GeV at the Υ( 4S ) resonance B (Υ( 4S ) → B ¯ B ) > 96 % ⇒ p CMS ( B ) + p CMS (¯ B ) + p CMS ( Beam ) = 0 ⇒ Clean event topologies (typically ∼ 10 tracks) t 1 tag side signal side K + π + π − t e − 2 π − π + t 3 t 4 t 5 Tag side and signal side of the full reconstruction Typical B factory event September 3rd, 2010 2/7 Daniel Zander – Full reconstruction

Improvements of the existing Full Reconstruction at Belle Addition of more B tag channels: Will increase the signal but . . . stage particles Especially high-multiplicity decays will tracks, K S , γ , π 0 increase the background drastically 1 ( s ) , D 0 and J /ψ D ± Maximum efficiency: No intermediate cuts, 2 usage of best B candidate, but . . . D ∗± ( s ) and D ∗ 0 3 Cuts after each stage are necessary due B ± and B 0 4 to finite computing power and storage capacity. NeuroBayes Multivariate analysis software utilizing a Neural Network The output of the Network can be interpreted as Bayesian probability Provides a well-discriminating variable for intermediate cuts September 3rd, 2010 3/7 Daniel Zander – Full reconstruction

Increase of Efficiency Choose appropriate cuts (depending on initial purity of the channel) Cut on the variable with good discrimination power X → A B C cut on NB ( A ) × NB ( B ) × NB ( C ) where NB ( N ) is the NeuroBayes output of particle N with NB ( N ) ǫ [ 0 . 0 , 1 . 0 ] Optimize these cuts depending on the decay channel September 3rd, 2010 4/7 Daniel Zander – Full reconstruction

D modes (new) K − π + D + K − π + π + D 0 → → D 0 K − π + π + π − D + K 0 S π + → → D 0 K − π + π 0 D + K 0 S π + π 0 → → D 0 π + π − D + K − π + π + π 0 → → D 0 π + π − π 0 D + K 0 S π + π + π − → → D 0 K 0 S π 0 D + K + K − π + → → D 0 K 0 S π + π − D + K + K − π + π 0 → → D 0 K 0 S π + π − π 0 D + ∗ D 0 π + → → D 0 K + K − D + ∗ D + π 0 → → D 0 K + K − K 0 → S D 0 ∗ D 0 π 0 D 0 ∗ D 0 γ → → D + D + K + K 0 K + K − π + π + π − → → S S S D + D + K + π + π − π + π + π − → → S S D + D + ∗ D + K + K − π + S γ → → S S D + K + K − π + π 0 e − e + J /ψ → → S D + K + K 0 S π + π − µ − µ + J /ψ → → S D + K − K 0 S π + π + → S September 3rd, 2010 5/7 Daniel Zander – Full reconstruction

B modes (new, limited D-modes in old full reconstruction) B + D 0 π + ¯ D − π + B 0 → → B + D 0 π + π 0 ¯ D − π + π 0 B 0 → → B + D 0 π + π + π − ¯ D − π + π + π − B 0 → → B + D + S ¯ ¯ D 0 B 0 D 0 π 0 → → ¯ B + D 0 ∗ π + D + B 0 → → S D − ¯ B + D 0 ∗ π + π 0 D ∗− π + B 0 → → ¯ B + D 0 ∗ π + π + π − D ∗− π + π 0 B 0 → → ¯ B + D 0 ∗ π + π + π − π 0 D ∗− π + π + π − B 0 → → S ¯ B + D + ∗ D ∗− π + π + π − π 0 D 0 B 0 → → S ¯ B + D + D + ∗ D 0 ∗ B 0 → → S D − S ¯ B + D + ∗ D + D 0 ∗ B 0 → → S D ∗− ¯ D + ∗ B + D 0 K + B 0 → → S D ∗− B + D − π + π + B 0 J /ψ K 0 → → S B + J /ψ K + B 0 J /ψ K + π − → → B + J /ψ K + π + π − B 0 J /ψ K 0 S π + π − → → B + J /ψ K + π 0 → B + J /ψ K 0 S π + → September 3rd, 2010 6/7 Daniel Zander – Full reconstruction

Conclusion Summary The full reconstruction is a powerful analysis tool at B factories Improvements were done with more sophisticated intermediate cuts and more B tag channels. Results Significant improvements could be achieved resulting in largely increased signal samples for many Belle analyses corresponding to several years of additional running time after 10 years of data taking Please visit the Full Reconstruction poster September 3rd, 2010 7/7 Daniel Zander – Full reconstruction