Speaker
Description
ABSTRACT
The lecture is devoted to the role of modern artificial intelligence (AI) methods in data analysis in the field of high-energy physics, astrophysics and particle physics. In these areas of scientific research, there is a rapid growth in the volume of data coming from particle detectors, telescopes and other complex measurement systems. Traditional methods of statistical analysis are becoming less sufficient for the effective processing and interpretation of such volumes and complexity of information, which makes AI methods - in particular machine learning (ML) and deep learning (DL) - an integral part of modern scientific methodology.
The lecture will provide an overview of the current state of AI applications in physics. Key cases will be considered, such as event classification in experiments at the Large Hadron Collider, signal extraction against noise in neutrino observatories, as well as pattern recognition in large arrays of astrophysical data. Special attention will be paid to neural network architectures that have already found application in physics, such as convolutional neural networks (CNN), recurrent networks (RNN), graph neural networks (GNN) and transformers. Current trends will also be covered: the implementation of AI in real time when entering data, explainable AI in tasks requiring physical interpretation, as well as methods for combating systematic errors and bias in model training. The lecture will conclude with an overview of the prospects and future challenges of using AI in high-energy physics.
BIO
Dr. Yury Shitov is a senior researcher at the Institute of Technical and Experimental Physics of the Czech Technical University (IEAP ČVUT, Prague), a specialist in neutrino physics, nuclear physics and elementary particle physics. He participates in a number of international experiments, including KM3NeT, DANSS, LEGEND, SuperNEMO and TGV. For over 30 years, he has been engaged in numerical modeling, statistical analysis and software development for processing and interpreting experimental data, including on distributed computing infrastructures.
In the last 5 years, he has paid special attention to the application of artificial intelligence methods in scientific research. In the KM3NeT experiment, he is developing AI tracking algorithms, creating a "super expert" - an intelligent system for supporting analysis and code development. He also gives theoretical and practical lectures on the use of AI in data analysis for students majoring in economics.
Author of over 120 scientific publications, Hirsch index - 25 (Scopus @ 09/06/2025).