Development and Explainability of Models for Machine-Learning-Based Reconstruction of Signals in Particle Detectors

Abstract

Machine learning methods are being introduced at all stages of data reconstruction and analysis in various high-energy physics experiments. We present the development and application of convolutional neural networks with modified autoencoder architecture for the reconstruction of the pulse arrival time and amplitude in individual scintillating crystals in electromagnetic calorimeters and other detectors. The network performance is discussed as well as the application of xAI methods for further investigation of the algorithm and improvement of the output accuracy.

This article was published in Particles, 2025, 8, 48, DOI: 10.3390/particles8020048

References in corpus (4)

• Occlusion Sensitivity Analysis with Augmentation Subspace Perturbation in Deep Feature Space
• Novel 3D Clustering Algorithm and Two Particle Separation with Tile HCAL
• Using Artificial Intelligence in the Reconstruction of Signals from the PADME Electromagnetic Calorimeter
• Applicability Evaluation of Selected xAI Methods for Machine Learning Algorithms for Signal Parameters Extraction