The school "Statistics for Physicists" was launched at Tomsk State University with the support of a megagrant from the Russian government. The participants of future experiments and developers of equipment for the Russian supercollider NICA master different methods of mathematical statistics, which will further help scientists to assess the accuracy of physical models when analyzing data.
The program of the school includes lectures and master classes on the following topics: Peculiarities of construction and testing of statistical hypotheses, Organization and launch of calculations on a computing cluster under HTCondor control, Consideration of statistical models applicable for the analysis of LHC data and others.
- Statistical methods are used not only to analyze the data obtained in experiments, but also to create the measurement equipment to get these data, explains Maxim Penzin, a teacher at the school and a member of the TSU Laboratory for High Energy Physics Data Analysis. - As part of the project implemented by TSU with the support of a megagrant, the researchers are modeling detectors and electronics for the supercollider. It is impossible to solve these problems without understanding the algorithms of mathematical statistics.
The researchers who will be directly involved in the experiments at the NICA collider need this tool more than ever. The big data that scientists obtain at megascience facilities are always "distorted" due to many factors.
At first, these are huge streams of information, which should be studied and analyzed. The methods of mathematical statistics help scientists to understand them in more details.
- It is not enough to get the data. You need to eliminate the unnecessary and analyze the information, preferably with a minimum amount of errors, says Maxim Penzin. - Statistical algorithms help to detect errors and take them into account in the analysis. The results of large-scale experiments are often reported in terms of confidence intervals of measured values or in terms of marginal error. For example, a certain particle was discovered with an accuracy of seven sigma (sigma is a standard deviation, characterizing the spread of a number of measurements). In other words, the methods of statistics help to determine the correctness of conclusions. The success of the whole experiment ultimately depends on it.
The school is online, since its participants are located in different parts of Russia.
At present, training of experts to work at the NICA supercollider is quite active. In March, the Joint Institute for Nuclear Research announced the start of the first session of the proton and heavy ion collider built in Dubna near Moscow. A beam of ions is expected to begin circulating in the collider ring in August. The first test data can be obtained in 2026.
