Options
Identification of heavy, energetic, hadronically decaying particles using machine-learning techniques
Date Issued
01-06-2020
Author(s)
Sirunyan, A. M.
Tumasyan, A.
Adam, W.
Ambrogi, F.
Bergauer, T.
Dragicevic, M.
Erö, J.
Del Valle, A. Escalante
Flechl, M.
Frühwirth, R.
Jeitler, M.
Krammer, N.
Krätschmer, I.
Liko, D.
Madlener, T.
Mikulec, I.
Rad, N.
Schieck, J.
Schöfbeck, R.
Spanring, M.
Waltenberger, W.
Wulz, C. E.
Zarucki, M.
Drugakov, V.
Mossolov, V.
Gonzalez, J. Suarez
Darwish, M. R.
De Wolf, E. A.
Di Croce, D. D.
Janssen, X.
Lelek, A.
Pieters, M.
Sfar, H. Rejeb
Van Haevermaet, H.
Van Mechelen, P.
Van Putte, S.
Van Remortel, N.
Blekman, F.
Bols, E. S.
Chhibra, S. S.
D'Hondt, J.
De Clercq, J. D.
Lontkovskyi, D.
Lowette, S.
Marchesini, I.
Moortgat, S.
Python, Q.
Skovpen, K.
Tavernier, S.
Van Doninck, W.
Van Mulders, P. V.
Beghin, D.
Bilin, B.
Clerbaux, B.
De Lentdecker, G. D.
Delannoy, H.
Dorney, B.
Favart, L.
Grebenyuk, A.
Kalsi, A. K.
Moureaux, L.
Popov, A.
Postiau, N.
Starling, E.
Thomas, L.
Velde, C. Vander
Vanlaer, P.
Vannerom, D.
Cornelis, T.
Dobur, D.
Khvastunov, I.
Niedziela, M.
Roskas, C.
Tytgat, M.
Verbeke, W.
Vermassen, B.
Vit, M.
Bondu, O.
Bruno, G.
Caputo, C.
David, P.
Delaere, C.
Delcourt, M.
Giammanco, A.
Lemaitre, V.
Prisciandaro, J.
Saggio, A.
Marono, M. Vidal
Vischia, P.
Zobec, J.
Alves, F. L.
Alves, G. A.
Silva, G. Correia
Hensel, C.
Moraes, A.
Teles, P. Rebello
Chagas, E. Belchior Batista Das
Carvalho, W.
Chinellato, J.
Coelho, E.
Abstract
Machine-learning (ML) techniques are explored to identify and classify hadronic decays of highly Lorentz-boosted W/Z/Higgs bosons and top quarks. Techniques without ML have also been evaluated and are included for comparison. The identification performances of a variety of algorithms are characterized in simulated events and directly compared with data. The algorithms are validated using proton-proton collision data at s = 13TeV, corresponding to an integrated luminosity of 35.9 fb-1. Systematic uncertainties are assessed by comparing the results obtained using simulation and collision data. The new techniques studied in this paper provide significant performance improvements over non-ML techniques, reducing the background rate by up to an order of magnitude at the same signal efficiency.
Volume
15