npsm 새물리 New Physics : Sae Mulli

pISSN 0374-4914 eISSN 2289-0041
Qrcode

Article

Research Paper

New Phys.: Sae Mulli 2019; 69: 813-817

Published online August 30, 2019 https://doi.org/10.3938/NPSM.69.813

Copyright © New Physics: Sae Mulli.

Research on Reactor Neutrino Event Selection by Using the Machine Learning Technique

Chang Dong SHIN1, Kyung Kwang JOO*1, Dong Ho MOON1, Myoung Youl PAC2, Junghwan GOH3

1Institute for Universe & Elementary Particles, Department of Physics, Chonnam National University, Gwangju 61186, Korea

2Institute of High Energy Physics, Dongshin University, Naju 58245, Korea
3Department of Physics, Kyung Hee University, Seoul 02447, Korea

Correspondence to:kkjoo@chonnam.ac.kr

Received: June 12, 2019; Revised: July 1, 2019; Accepted: July 2, 2019

This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

For the next-generation massive neutrino experiments, selecting a signal in the background events is very important. To do this, we investigated the results of applying a machine learning technique to the selection of neutrino signals. The neutrino signal after inverse beta decay and the background events in a gadolinium-loaded liquid scintillation detector were reproduced by using Monte Carlo simulations. The inverse beta decay process is well-known and has relatively high statistical quantities for this simulation. In this study, an efficiency of signal selection through machine learning was obtained, and in this paper several results are briefly described. Finally, the machine learning technique is expected to become an important tool for use in the next-generation neutrino experiment.

Keywords: Machine learning, Neutrino, Liquid scintillator, Neutrino mixing angle, Next generation neutrino detector

차세대 대형 중성미자 검출 실험에서 배경사건과 원하는 신호를 효율적으로 선별하는 것은 매우 중요하다. 이를 위해서 현재 유용하게 사용되고 있는 분석 기술의 하나인 기계학습을 사용하여 중성미자 신호 선별에 적용하였을 때의 결과를 살펴보고자 한다. 이를 위해 비교적 특징이 잘 알려지고, 상대적으로 통계량이 높은 원자로 중성미자의 역베타 붕괴 반응 이후 신호와 배경사건들을 몬테카를로 시뮬레이션을 통하여 재현하고, 기계학습을 통한 신호선별 효율을 확인하였다. 최종적으로는 향후 차세대 중성미자 실험에서 중요한 도구로 사용될 수 있을 것으로 기대한다.

Keywords: 기계학습, 중성미자, 액체섬광검출용액, 진동변환상수, 차세대 중성미자 검출기

Figures

Fig. 1. Background rejection (a) and signal efficiency (b) by machine learning. The red line is the result using all variables and the green line is the result using only $\Delta T$ and $\Delta R$ without energy variables. Both results show that background can be removed efficiently by machine learning.

Stats or Metrics

Share this article on :

Related articles in NPSM