中国物理学会高能物理分会第十三届全国粒子物理学术会议（2021）

Large-volume liquid scintillator detector with ultra-low background levels have been widely used to study neutrino physics and search for dark matter. The ability to accurately reconstruct particle interaction events is of great importance for the success of the experiment. The signal collected by PMTs is used for estimation of the vertex and the energy of neutrino and background particle interactions. In this work we present several machine learning approaches applied to the vertex and the energy reconstruction. Multiple models and architectures were compared and studied, including Boosted Decision Trees (BDT), Deep Neural Networks (DNN), a few kinds of Convolution Neural Networks (CNN), based on ResNet and VGG, and a Graph Neural Network based on DeepSphere. The models of BDT and DNN are trained with aggregated information, pre-calculated from PMT signal, while the others are trained with PMT-wise measured information from PMTs.