electron-hadron separation by neural-network yonsei university
TRANSCRIPT
![Page 1: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/1.jpg)
Electron-hadron separation
by Neural-network
Yonsei university
![Page 3: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/3.jpg)
TRD(Transition Radiation Detector)
• |η|<0.9, 45°<θ<135°
• 18 supermodules in Φ sector
• 6 Radial layers
• 5 z-longitudinal stack
total 540 chambers
750m² active area
28m³ of gas
• In total 1.18 million read-out
channels
FROM C. Adler -- Hadron Collider Physics, Les Diablerets, 4-9 July 2005
![Page 4: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/4.jpg)
TRD(Transition Radiation Detector)
(What is transition radiation?)“Transition radiation is emitted whenever a charged particle crosses an interface between two media with different dielectric functions.”
- L.Durand, Phys. Rev. D 11, 89(1975)
• Predicted : Ginzburg & Frank, 1946
• Observed : Goldsmith & Jelly, 1959(optical)
• It’s sizeable(X-rays) for relativistic particles.
FROM A. Andronic -- an overview for students, GSI, Feb. 6th, 2007
![Page 5: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/5.jpg)
TRD working principle
Cut through one side of
a TRD readout chamberFROM P. Shukla -- ICPA-QGP'05, Kolkata, 8-12
The total energy loss by TR for charged particle : depends on its Lorentz factor γ = E / mc2
Usefullness of TR : For discrimination the electron from hadons in the momentum range between 1GeV/c and 100GeV/c.
![Page 6: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/6.jpg)
1 Event signals of Electron and Pion
FROM P. Shukla -- ICPA-QGP'05, Kolkata, 8-12 February 2005
![Page 7: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/7.jpg)
e/π-Separation
Likelihood to be electrons
PP
PL
e
e
N
ii
N
iie
XPP
eXPP
1
1
),(
)(
32001000
32001800
32001000
60001800
32001800
60001800
![Page 8: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/8.jpg)
e/π-Separation-Classical Methods
Distributions of the likelihood for electrons and pions of 2GeV/c, obtained from the total energydeposit.
Pion efficiency as a function of electron efficiency for a 6 layer likelihood on total energy, for momentum of 2GeV/c. The values corresponding to measured data are comparedsimulations.
![Page 9: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/9.jpg)
e/π-Separation-Classical Methods
Pion efficiency as a function of number of layers for a momentum of 2GeV/c
![Page 10: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/10.jpg)
Other method?
Neural-network - A simple Modeling of the biological neuron system - Being used in various fields for data
analysis and classification - Examples : Image analysis Financial movement’s prediction Sales forecast In particle physics
![Page 11: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/11.jpg)
Modeling of neuron
], ... ,,[
], ... ,,[
321
321
n
n
wwwwW
xxxxX
UNIT y
Output
x1
xn
Iuputs
Weight of connection
![Page 12: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/12.jpg)
NN-Example
], ... ,,[
], ... ,,[
321
321
n
n
wwwwW
xxxxX
NET
f
3
1Neuron
Training sample( x1=1, x2=2, x3=3, y=1 )
![Page 13: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/13.jpg)
NET
f
3
1
Neuron
!! But we want the output to be 1Change the W vector by some rules
1.11 w
5.03 w
4.02 w
1)(
4.3
NETfy
XWNET T
1y
NN-ExampleTraining sample( x1=1, x2=2, x3=3, y=1 )
![Page 14: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/14.jpg)
Adc[1]
PionOr
electron
e/π-Separation-Neural network
Adc[2]
Adc[3]
Adc[20]
![Page 15: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/15.jpg)
Comparison of methods
![Page 16: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/16.jpg)
Summary
• The suppression factor was improved by about a factor of 3 by Neural net when compared with classical method.
• Need to be done – Application to the full simulation sample– Refined Neural Network Other NN structures must be considered
![Page 17: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/17.jpg)
`
![Page 18: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/18.jpg)
![Page 19: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/19.jpg)
![Page 20: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/20.jpg)
![Page 21: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/21.jpg)
Modeling of Neuron
x Σ f(NET) y
InputCombinatio
nfunction
OutputActivationfunction
NET
![Page 22: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/22.jpg)
Ionization energy loss for muons and pions and protons on a variety of materials.
Bethe-Bloch formula
225
2222
2
2
101.5
])2
[ln(
MeVcmD
I
cmnDq
dx
dE ee
Ionization energy loss
![Page 23: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/23.jpg)
Momentum dependence of the most probable and the average values of the energy loss for pions and electrons. The symbols are measurements, the lines are calculations
Ionization energy loss
![Page 24: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/24.jpg)
Signals of Electron and Pion
Many Events
Averaged!!
![Page 25: Electron-hadron separation by Neural-network Yonsei university](https://reader035.vdocuments.site/reader035/viewer/2022062409/5697bf721a28abf838c7ec5e/html5/thumbnails/25.jpg)
Mean signal of Electron and Pion