HelingZhuSandbox

This is Heling Zhu's sandbox to store my studies in ATLAS Experiment.

Physics Analysis

Resonance search in low-mass 4-muon final states

Run-1 + data 2015/16 analysis

• In Run1, the local significance is 5.4 without trigger, and 3.7 with EF_3mu4T trigger.
• No clear evidence of signal in data 2015+2016.
• Run-1+2015/16 Supporting note
• Unblind approval meeting for data 2015/16 by Xiangyang

Data 2017 analysis

• 13TeV default trigger: HLT_3mu4.
• Data 2017 Supporting note
• Unblind approval meeting for data 2017 by Bing Li
• Analysis Summary of unblind 2017 data in EB meeting by Heling

New baseline selection

• Use reprocessed 8TeV data and 13TeV data from the years of 2015-2017
• New supporting note preparing for 2018 unblinding
• Talk in BLS plenary during ATLAS p&p week in July 2019. by Heling
• Summary of latest (last) results in 'baseline' analysis in USTC group meeting. by Heling

Measurements of VBSVV semi-leptonics process

• TWiki page
• Supporting note
• Paper published in Physics reivew D (PLD)
• Paper approval meeting by Heling
• Common Package for this analysis in reco-level: CxAODReader_VBSVV
• Package for truth level study (fiducial space): xAODTruth_VBSVV
• Tool for this analysis: To run systematic inputs with large variations, usually many jobs failed in lxbatch. One can use this tool to resubmit failed jobs quickly.

HZZ analysis

Off-shell Higgs analysis in data 2015+2016

• Paper has been published in Physics Letter B.
• Analysis Summary in 2018 HZZ workshop by Luigi
• Talk in 2018 APS April meeting by Heling
• Tool for this analysis: Plot template for Off-shell paper

EFT interpretation in unfolding Inclusive m4l in data 2015+2016

• Twiki page
• Paper has been published in JHEP.
• Paper presentation in ATLAS weekly by Heling
• Paper approval meeting by Rongkun
• Package for this analysis: It is not a package for an overall analysis, just the package to run EFT interpretation in both Asymptotic and MC toy method, and submitting condor for MC toy method.

Commom H4l analysis

• TWiki page of H4l overview
• Derivations for Higgs group
• Results of lepton isolation studies in MCP meeting during ATLAS p&p week in Jan. 2019. by Heling

High mass Heavy resonance search in full run-2

• Supporting note
• Paper in arXiv, the paper has been submitted to EPJC
• Talk in Higgs plenary during ATLAS p&p week in July 2019. by Heling
• Paper Approval Meeting by Heling
• Packages for this analysis
  • To make HZZ Minitrees: HZZAnalRun2Code)
  • To make workspace for 4l and llvv channels: HZZWorkspace; TWiki of HiggsZZRunIIWorkspaces
  • Scripts for some simple fitting: CheckWorkspace
  • To make combined workspace: TWiki of WorkspaceCombiner

VBSZZ analysis

HL-HLC VBSZZ

• Physics results are public as ATLAS PUBnote
• The results are included into Yellow Report public in arXiv.
• Standard Model Approval Meeting by Heling
• Working on some followups for HGTD. Talk in Upgrade meeting during ATLAS p&p week in Jan. 2019. by Heling

Observation of VBSZZ with full run-2 datasets

• Conf note for EPS
• Paper in arXiv, the paper has been submitted to Nature Physics
• Paper approval meeting. by Shuzhou
• Talk in Corfu 2019. by Heling

Service Works

Muon DQ Offline Monitoring Shift

• Website to book a shift.
• Summary Tutorial by Rongkun
• Indico page of shift training
• Shift in 2018: April 26 (Thu) - May 2nd (Wed), report in DQ meeting

The development of the tests and evaluations of the LAr Trigger Digitizer Boards (LTDB) of the ATLAS LAr Calorimeter Phase-1 Upgrade

• The LTDB boards we have studied and their related documents (by Heling):
  • 320-ch pre-prototype
  • 64-ch prototype
  • 320-ch prototype
  • 320-ch pre-production: Board #1, Board #2, Mapping and test status
• Presentation: LAr Week of Mar. 2017, FE Electronics WG for Phase-1 upgrade (May 2017), System test and demonstrator integration weekly (Aug. 2017) by Heling

• Software development works at CERN:
  • Mapping between LTDB and LATOME data taking system.
  • Talk in LAr week in March 2019. by Heling
  • Talk with problem (in 2019 LAr week talk) fixed. by Heling

Other meetings and discussion

Analysis discussion with Hong Ma

• Before Oct. 2018: every Thursday 5pm-6pm @ BNL
• After Oct. 2018: Thursday 7pm-8pm biweekly @ CERN
• Sets of my slides

Deep learning

Parameterized Deep Neural Network for Z prime search

• Reference in arXiv
• My codes in GitHub, based on Keras, Tersorflow backend

COVID-19 Medical paper Named-entity recognition (NER)

• Fine-tuning bases on pretraining model from BERT
• Codes use GPU on Google Colab.

Optimization of RPC TOF (time-of-flight) measurement resolution

• Two models: LSTM, CNN, training with GPU in Colab

Coursera Licenses and Certifications

• Applied Machine Learning in Python, University of Michigan, Credential ID 8LM6MHB3YNLR
• Machine Learning, Stanford University, Credential ID Z352QHAY5226

-- HelingZhu - 2018-04-23