|国家预印本平台
| 注册
首页|Machine Learning methods for event classification and vertex reconstruction of the 12C + 12C reaction with the MATE-TPC

Machine Learning methods for event classification and vertex reconstruction of the 12C + 12C reaction with the MATE-TPC

Minghui Zhang Xiaobin Li Jie Chen Ningtao Zhang Fenhua Lu Junrui Ma Jiazhen Yan Wanqin Tu Xiaodong Tang Bingshui Gao Chengui Lu Zhichao Zhang Jinlong Zhang Weiping Liu

✕
中国科学院科技论文预发布平台_logo中国科学院科技论文预发布平台

Machine Learning methods for event classification and vertex reconstruction of the 12C + 12C reaction with the MATE-TPC

Machine Learning methods for event classification and vertex reconstruction of the 12C + 12C reaction with the MATE-TPC

Minghui Zhang 1Xiaobin Li 1Jie Chen 1Ningtao Zhang 2Fenhua Lu 1Junrui Ma 1Jiazhen Yan 3Wanqin Tu 1Xiaodong Tang 2Bingshui Gao 2Chengui Lu 2Zhichao Zhang 2Jinlong Zhang 2Weiping Liu1

作者信息

  • 1. Southern University of Science and Technolog
  • 2. Institute of Modern Physics Chinese Academy of Sciences
  • 3. Peking University
  • 折叠

摘要

In modern nuclear physics experiments, identifying the events of interest is challenging for nuclear reactionstudies with the active target Time Projection Chamber (TPC). In this work, machine learning techniques areemployed to analyze the complex data of 12C + 12C fusion reaction from a TPC named MATE (multi-purposeactive-target time projection chamber for nuclear experiments). Specifically, we successfully applied Resid-ual Neural Network (ResNet-50, ResNet-34 and ResNet-18) and Visual Geometry Group (VGG-19) to classifyelastic scattering and fusion reaction events from the 12C + 12C reaction. The classification results of the fourmodels are nearly identical, with accuracies of approximately 97% for the simulated data and 90% for the exper-imental data. Moreover, these approaches successfully identify some events that are misclassified by traditionalmethods. These models are also applied to classify events from different fusion reaction channels, their classi-fication accuracies are approximately 95% on the simulated data. In addition, a Convolutional Neural Network(CNN) model is developed to reconstruct the reaction vertex, providing an alternative strategy for vertex recon-struction. These results indicate that machine learning techniques can effectively classify the reaction events ofdifferent reaction channels and reconstruct the reaction vertex, thereby paving the way for future analyses ofcomplex nuclear reaction data.

Abstract

In modern nuclear physics experiments, identifying the events of interest is challenging for nuclear reactionstudies with the active target Time Projection Chamber (TPC). In this work, machine learning techniques areemployed to analyze the complex data of 12C + 12C fusion reaction from a TPC named MATE (multi-purposeactive-target time projection chamber for nuclear experiments). Specifically, we successfully applied Resid-ual Neural Network (ResNet-50, ResNet-34 and ResNet-18) and Visual Geometry Group (VGG-19) to classifyelastic scattering and fusion reaction events from the 12C + 12C reaction. The classification results of the fourmodels are nearly identical, with accuracies of approximately 97% for the simulated data and 90% for the exper-imental data. Moreover, these approaches successfully identify some events that are misclassified by traditionalmethods. These models are also applied to classify events from different fusion reaction channels, their classi-fication accuracies are approximately 95% on the simulated data. In addition, a Convolutional Neural Network(CNN) model is developed to reconstruct the reaction vertex, providing an alternative strategy for vertex recon-struction. These results indicate that machine learning techniques can effectively classify the reaction events ofdifferent reaction channels and reconstruct the reaction vertex, thereby paving the way for future analyses ofcomplex nuclear reaction data.

关键词

MATE-TPC/machine learning/event classification/vertex reconstruction/Residual Neural Network

Key words

MATE-TPC/machine learning/event classification/vertex reconstruction/Residual Neural Network

引用本文复制引用

Minghui Zhang,Xiaobin Li,Jie Chen,Ningtao Zhang,Fenhua Lu,Junrui Ma,Jiazhen Yan,Wanqin Tu,Xiaodong Tang,Bingshui Gao,Chengui Lu,Zhichao Zhang,Jinlong Zhang,Weiping Liu.Machine Learning methods for event classification and vertex reconstruction of the 12C + 12C reaction with the MATE-TPC[EB/OL].(2026-05-09)[2026-05-13].https://chinaxiv.org/abs/202605.00079.

学科分类

计算技术、计算机技术

评论

首发时间： 2026-05-09
下载量：0
|
点击量：17
段落导航相关论文