Neural network for mass reconstruction of resonance particle with missing energy

Zhang Zi-Ping

Nuclear Science and Techniques

pp.65-68

Published in print 01 May 1996

19000

Neural Network can be designed to reconstruct the mass of resonance particle with large missing energy. Taking the Higgs particle search through decaying channel H⁰→τ⁺τ^- →eµx and H⁰→W⁺W^-(ZZ)→llvv at LHC collider ( $\sqrt{s}$ TeV) as examples, neural network correctly reconstructs its mass with right peak position and better width than conventional method. The network also possesses the capability of suppressing background events. This kind of neural network can be widely used in new particle search and precise mass measurement of resonance particle.

Neural networkMass reconstructionResonance particlesHiggs search

References

Bellantoni L, Conway J S, Jacobsen J E et al., CERN-PPE/91-80, 1991

Innoc ente V, Wang Y F, Zhang Z P. Nucl Instr and Meth, 1992; A323: 647

Bortolotto C, De Angelis A, Lanceri L. Nucl Instr and Meth, 1991; A306: 459

Lönnblad L, Peterson C, Rögnvaldsson T. Phys Lett, 1992; B278: 181

Joswig M, Kolanoski H, Kublun S et al., DESY 93-167, preprint

Lella L D. CERN 90-10/ECFA 90-133, 1990; II: 530

Bos K, Anselmo F, Eijk B. CERN 90-10/ECFA 90-133: 538

Kunszt Z, Zwirner F. CERN 90-10/ECFA 90-133: 578

Sjöstrand T. PYTHIA5.6 and JETSET7.3 physics and mannual, CERN-TH. 6488/92, 1992

Lönnblad L, Peterson C, Rögnvaldsson T. Nucl Phys, 1991; B349: 675