The Study of Jet Tagging Algorithm Based on DeepSets at CEPC

doi:10.11871/jfdc.issn.2096-742X.2024.03.012

Abstract

Abstract:

[Purpose] After the existence of the Higgs boson was confirmed, precision measurements of its properties and searches for new physics beyond the Standard Model become the focus of the field of high energy physics in the future, and better detection of the jet can effectively improve the precision and sensitivity. [Method] In this study, a newly proposed DeepSets-based deep learning algorithm, the Particle Flow network, and a collection of full simulation data sample of jets generated by the hadronic decay of the Z boson are used to study the tagging algorithm of heavy flavor jets. [Results] Compared with the CEPC baseline algorithm, the performance of the flavor tagging is improved by about 6%, i.e., the average accuracy is increased from 80% to 85%. [Conclusion] This study shows that deep learning has great potential in high-energy physics experiments, and has wide application scenarios in particle reconstruction, physics analysis, and other aspects. However, more in-depth research on deep learning methods is needed to enhance the credibility and robustness.

Key words: deep learning, particle flow network, jet flavor tagging

LIAO Libo, WANG Shudong, SONG Weimin, ZHANG Zhaoling, LI Gang, HUANG Yongsheng. The Study of Jet Tagging Algorithm Based on DeepSets at CEPC[J]. Frontiers of Data and Computing, 2024, 6(3): 108-115, https://cstr.cn/32002.14.jfdc.CN10-1649/TP.2024.03.012.

Figures/Tables 8

Fig. 1

Table 1

Fig.2

Table 2

Input variables for PFN"

输入变量	描述
$l o g E$	粒子能量的对数
$c o s θ$	粒子极角的余弦
$?$	粒子的方位角
PID	经鉴别后的粒子ID
$D 0$	粒子次级顶点到束流的垂直距离
$Z 0$	粒子次级顶点到束流对撞点的水平距离
$l o g M$	粒子动量长度的对数
$s i g D 0$	粒子次级顶点到束流的垂直距离比上该项的误差
$s i g Z 0$	粒子次级顶点到束流对撞点的水平距离比上该项的误差
$δ R$	粒子与喷注之间的角距离
$? · s i n θ$	粒子方位角与极角正弦的乘积
$δ θ$	粒子与喷注之间的极角差
$δ ?$	粒子与喷注之间的方位角差
$q$	粒子的电荷

Table 2

Table 3

Fig.3

Fig.4

Table 4

The production of efficiency and purity at different tagging efficiency"

标记效率 $ε S$	50%	90%	95%
$b b$	47.5%	80.1%	79.7%
$c c$	47.5%	74.7%	69.4%
$o o$	48.5%	72.9%	68.4%

Table 4

References 27

[1]	CMS C. Observation of a new boson at a mass of 125 GeV with the CMS experiment at the LHC[J]. Physics Letters B, 2012, 716(2012): 30-61.
[2]	ATLAS C. Observation of a new particle in the search for the Standard Model Higgs boson with the ATLAS detector at the LHC[J]. Physics Letters B, 2012, 716(2012): 1-29.
[3]	CHEN T Q, GUESTRIN C. Xgboost: A scalable tree boosting system[C]. Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining. ACM, 2016: 785-794.
[4]	SUEHARA T, TANABE T. LCFIPlus: A framework for jet analysis in linear collider studies[J]. Nucl. Instrum. Meth, 2016, A808(2016): 109-116.
[5]	BALDI T, BAUER K, ENG C, et al. Jet substructure classification in high-energy physics with deep neural networks[J]. Physical Review D, 2016, 93(9): 094034.
[6]	BARNARD J, DAWE E N, DOLAN M J, et al. Parton shower uncertainties in jet substructure analyses with deep neural networks[J]. Physical Review D, 2017, 95(1): 014018.
[7]	GUEST D, COLLADO J, BALDI P, et al. Jet flavor classification in high-energy physics with deep neural networks[J]. Physical Review D, 2016, 94(11): 112002.
[8]	COGAN J, KAGAN K, STRAUSS E, et al. Jet-images: Computer vision inspired techniques for jet tagging[J]. Journal of High Energy Physics, 2015, 02: 118.
[9]	LOUPPE G, CHO K, BECOT C, et al. QCD aware recursive neural networks for jet physics[J]. Journal of High Energy Physics, 2019, 01: 057.
[10]	ZAHEER M, KOTTUR S, RAVANBAKHSH S, et al. Deep sets[C]. 2017, 3394-3404.
[11]	KOMISKE P T, METODIEV E M, THALER J. Energy flow networks: Deep sets for particle jets[J]. Journal of High Energy Physics, 2019, 01: 121.
[12]	NAIR V, HINTON G E. Rectified linear units improve restricted boltzmann machines[C]. Proceedings of the 27th International Conference on International Conference on Machine Learning, 2010, 807-814.
[13]	HE K M, ZHANG X Y, REN S P, et al. Delving deep into rectifiers: Surpassing human-level performance on imagenet classification[C]. 2015 IEEE International Conference on Computer Vision, 2015: 1026-1034.
[14]	GALLICCHIO J, SCHWARTZ M D. Quark and gluon tagging at the LHC[J]. Physical Review Letter, 2011, 107: 172001.
[15]	GALLICCHIO J, SCHWARTZ M D. Quark and gluon jet substructure[J]. Journal of High Energy Physics, 2013, 04: 090.
[16]	ATLAS COLLABORATION. Light-quark and gluon jet discrimination in pp collisions at √s = 7 TeV with the ATLAS detector[J]. The European Physical Journal C, 2014, 74(8): 3023.
[17]	GRAS P, HOCHE S, KAR D, et al. Systematics of quark/gluon tagging[J]. Journal of High Energy Physics, 2017, 7: 091.
[18]	FRIEDMAN J H. Greedy function approximation: a gradient boosting machine[J]. Annals of statistics, 2001, 29(5): 1189-1232.
[19]	ZHOU Z H, FENG J. Deep forest: Towards[J/OL]. 2020-6-6 [2023-9-25]. https://arxiv.org/abs/1702.08835.
[20]	KILIAN W, OHL T, REUTER J. WHIZARD: Simulating multi-particle processes at LHC and ILC[J]. The European Physical Journal C, 2011, 71: 1742.
[21]	MORETTI M, OHL T, REUTER J. O'Mega: An optimizing matrix element generator[J/OL]. 2001-2-15 [2023-9-25]. https://arxiv.org/abs/hep-ph/0102195.
[22]	THE CEPC STUDY GROUP. CEPC conceptual design report: Volume 2-Physics & Detector[J/OL]. 2018-11-23 [2023-9-25]. https://arxiv.org/abs/1811.10545.
[23]	RUAN M Q, VIDEAU H. Arbor, a new approach of the particle flow algorithm[J/OL]. 2016 [2023-9-25]. https://arxiv.org/abs/1403.4784v2.
[24]	QU H L, GOUSKOS L. ParticleNet: Jet tagging via particle clouds[J]. Physical Review D, 2020, 101(5): 056019.
[25]	QU H L, LI C Q, QIAN S T. Particle transformer for jet tagging[J/OL]. 2022 [2023-9-25]. https://arxiv.org/abs/2202.03772.
[26]	GONG S Q, MENG Q, ZHANG J. et al. An efficient Lorentz equivariant graph neural network for jet tagging[J]. Journal of High Energy Physics, 2022, 7: 030.
[27]	DREYER F A, and QU H L. Jet tagging in the Lund plane with graph networks[J]. Journal of High Energy Physics, 2021, 03: 052.