High Energy Physics Data Analysis Laboratory

• The Laboratory has collected, reconstructed, and final analysis of detector data from the Large Hadron Collider.
• We have have found a Higgs boson in the main channel of decay two bottom quarks. The discovery of this decay channel has great importance for proving the Standard model of elementary particle and gives a significant impulse for continuing investigation.
• Our team has conducted research to search for a channel of Higgs boson decay into two muons. This channel has very low relative width of decay and is not available for observation due to small statistical data sample. Nevertheless, the found upper bound of scattering cross section multiplied by decay rate reaches the value predicted by the Standard model.

Implemented results of research:

• We have created electronics for testing VMM3 chips and developed methods for their complex testing for new detectors. Apart from that, a group in the ATLAS experiment is responsible for operations of cathode strip detectors of the ATLAS.
• Our team has developed specialized software that ensures stable and fault-free of gauge constants of all the channels of VMM into databases including necessary tools for validation and analysis of saved data.
• We have developed a graphical interface and a web interface simultaneously ensuring security of network resources. Obtained results and accumulated experience can be used in modern microelectronics for creating and tasting new electronic systems.
• Algorithms and methods for analysis of big data developed on the grounds of the Laboratory are in high demand in high tech applied research fields such as materials science, modern engineering analysis (CAD/CAE systems) for supporting R&D and information systems operating on big data and super big data.

Education and career development:

• 3 students and 3 postgraduates have been hired by the Laboratory to conduct their dissertations on the topic of the project under the auspices of the leading scientist and members of the scientific team of the Laboratory.
• We have opened the «Elementary particle physics» masters program. The goal of the program is to train professionals for working at major world class research projects, including experiments of the European Organization for Nuclear Research (CERN) in Switzerland and the Joint Institute for Nuclear Research in Dubna (Russia). Masters students will learn to analyze data obtained in complex physical experiments.
• The Laboratory has conducted a series of lectures in experimental high energy physics as a part of training for students, postgraduates, young employees for analyzing ATLAS data imitating processes accompanying passage of charged particles in the ATLAS detector and their visualizations by studying technologies of detecting ionizing radiation in ATLAS at the European Organization for Nuclear Research (CERN).
• We have conducted scientific seminars with participation of invited scientists as well as employees of scientific and educational foundations of Tomsk.

Organizational and structural changes:

• We have developed a concept of a computation cluster in a configuration that meets the requirements concerning amount of data generated by the ATLAS experiment in head-on proton collisions, speed of its processing as well as generating data using the Monte Carlo method. According to the developed computation cluster concept in 2018 we installed and launched the first lot of computational equipment comprising 30% of projected capacity. As a caching system we have configured the Squid software package that acts as a proxy server for HTTP, FTP, HTTPS protocols. This package ensures compatibility with the majority of the most important Internet protocols as well as with the majority of widely used operating systems.
• The Tomsk State University grants access to the SKIF-Cyberia computer cluster connected with the Kurchatov Institute and the European Organization for Nuclear Research (CERN). The Laboratory has direct connection to resources of the Tomsk State University through the University's local network with 40 Gbit/sec bandwidth.
• In the coming years computational capabilities of the Laboratory will increase, the engineering infrastructure is now ready for accommodating up to 10 PB of storage space and 100 teraflops of operational performance.

Other results:

• We have modernized algorithms of research of characteristics of the Higgs boson. The proposed new approach that uses the Docker container technologies allows to ensure replicability of results obtained on the basis of machine learning algorithms. On the recommendation of the ATLAS machine learning forum, developed materials have received the official Tutorials status and they can be currently be accessed freely to engage physicists in using state-of-the-art data processing tools. Tutorials introduce ATLAS physicists to implementation of the TMVA algorithm based on the AdaBoost algorithm from the popular SCIKIT_LEARN machine learning library, as well as to applications of the more modern and productive XGBoost algorithm. We have demonstrated the potential of newest machine learning methods for increasing importance of discovery of Higgs decays.

Collaborations:

• Brookhaven National Laboratory (USA): joint research and scientific publications
• High Energy Accelerator Research Organization – KEK (Japan): joint research, scientific events, scientific publications
• Joint Institute for Nuclear Research (Russia): joint research, scientific events, joint education research, academic exchanges, scientific publications