Towards release of the LCFI Vertex Package

1. Towards release of the LCFI Vertex Package • Flavour tag • Quark sign selection / vertex charge reconstruction • What remains to be done before the software release

2. tanh (MPt / 5 GeV) b c joint probability uds b c Flavour tag • Vertex package will provide flavour tag procedure developed by R. Hawkings et al • (LC-PHSM-2000-021) and recently used by K. Desch / Th. Kuhl as default • NN-input variables used: • if secondary vertex found: MPt , momentum of secondary vertex, and its decay length and decay length significance • if only primary vertex found: momentum and impact parameter significance in R-f and z for the two most-significant tracks in the jet • in both cases: joint probability in R-f and z (estimator of • probability for all tracks to originate from primary vertex) • will be flexible enough to permit user further tuning of the input variables for the neural net, • and of the NN-architecture (number and type of nodes) and training algorithm

3. Initial checks: flavour tag at the Z-peak • use SGV and FORTRAN ZVTOP with neural net software developed at • Bristol (Dave Bailey) – BaBar, looking for NN-code to include in their analysis software, • chose this NN implementation because of its user-friendliness • look at q qbar sample (all flavours except top) at the Z peak; this consists of: • train separate nets for jets with 1, 2 and 3 ZVTOP vertices found; • for each category, train three types of net: b-net and c-net to identify b- and c-quarks, • respectively and bc-net to identify c-quarks with background only from b-jets • plot flavour tag purity vs efficiency; angular cut 30o < qthrust < 150o

4. K Desch/ Th Kuhl b c (b bkgr) c Flavour tag purity vs efficiency at the Z-peak • better performance than obtained previously(TESLA-TDR, Desch/Kuhl): • high b-jet purity up to larger efficiencies; • c-purity for c with only b-background higher; also improvement for c-tag • reason might be masking of K-shorts and Lambdas in current study, actual efficiency • for doing this has still to be evaluated

5. uds b c c b uds Flavour composition of sample at different energies left: contributions of b, c and uds jets in generated sample, right: require > 1 ZVTOP vertex fractions of b- and c-jets become more similar at higher energies in that respect, b-tag becomes more challenging; increase in average decay length makes vertex finding easier

6. c (b bkgr) b c Tagging performance at ECM = 500 GeV left: momentum and decay length significance of secondary vertex and momentum of most and of second-most significant track in jet scaled with jet energy • b-tag worse, c (b-bkgr) tag improved at 500 GeV compared to Z-peak, consistent with change • in contributions of b- and c-jets to sample • comparison with previous result (right) not fully possible (K-short and Lambda-masking)

7. Vertex charge reconstruction • in the 40% of cases where b quark hadronises to charged B-hadron quark sign can • be determined by vertex charge • need to find all stable tracks from • B decay chain: • define seed axis • cut on L/D (normalised distance • between IP and projection of track POCA • onto seed axis) • tracks that form vertices other than IP • are assigned regardless of their L/D • probability of mis-reconstructing vertex charge is small for both charged and neutral cases • neutral vertices require ‘charge dipole’ procedure from SLD still to be developed for ILC, • relying on ghost track / ZVKIN

8. purity value obtained using only MPt for flavour tag Vertex charge for inclusive sample at the Z peak • look at purity and efficiency of finding vertex charge other than 0, plotted vs b-tag efficiency 0.5 efficiency_pm 0 0.5 • previously used pure b bbar sample and cut on MPt for ‘flavour-tag’, typical cut value 2 GeV • corresponding to ~ 65 % efficiency; grey line: average over polar angle range considered •  good agreement with previous results

9. Completion of first release of ‘vertex package’ • code development: • coding: complete ghost track algorithm, add C++-based calculation of NN input variables • interface NN code from Bristol, with flexible input (default: Hawkings procedure) • interface vertex package to SGV and to LCIO • in parallel: document • testing: • ZVRES: detailed comparison to FORTRAN version and to MC (in progress) • ZVKIN testing, e.g. comparison with SLD performance at 45 GeV jet energy • check flavour tag and vertex charge performance at each stage • after release: • evaluation of differences between SGV and full simulation and reconstruction framework • study possible improvements in flavour tag and vertex charge, e.g. from ghost track Many interesting topics to work on – further help with exploring these most welcome

10. Recent and future WP1-related talks LCWS 9 – 14 March 2006, Bangalore: S. Worm, ‘LCFI Status Report: Physics and Mechanics for ILC Vertex Detector’ ILC Software and Physics meeting, 4 – 6 April 2006, Cambridge: S. Hillert, ‘Overview of LCFI Vertex Package’ B. Jeffery, ‘Towards a C++ based version of ZVTOP’ ‘Linear Collider: Detector and Physics’ Forum, 6 – 7 May, Cosener’s House: T. Greenshaw, ‘LCFI Physics Studies’

11. Additional Material

12. c (b bkgr) b c c (b bkgr) b c Tagging performance at ECM = 500 GeV NB difference in range: below: ~0.3 – 1 right: ~ 0.0 - 1 Inputs scaled Inputs not scaled