CLUSTER REGRESSION ANALYSIS TO DETERMINE THE IMPACT OF GDP ON VOLUNTARY HEALTH INSURRANCE PREMIUS

Increasingly, the policy of the Russian Federation is aimed at the development of healthcare, and spending on the healthcare industry is becoming a serious problem for the financial stability of national health systems, even in highincome countries. Voluntary medical insurance, as one of the forms of medical insurance, contributes to the formation of all insurance premiums, thereby this form can be one of the indirect factors of socioeconomic development of Russia. The paper hypothesizes the existence of a relationship between the level of development of the welfare of the country and the volume of the premium fund of voluntary health insurance. A study was conducted on the relationship between voluntary health insurance and GDP, as one of the key indicators of the country's development, through correlation and regression analysis.

MEASUREMENT TRUSTWORTHIINESS AND NEW TECHNOLOGIES

The analysis of tendencies related to provision of measurement trustworthiness has made it possible to propose a methodology for automatic checking metrological health of measuring instruments and multichannel measuring systems in the course of their operation. This methodology is called «metrological selfcheck ». The selfcheck enables calibration/verification intervals to be significantly increased. In case of virtual test and artificial intelligent application, automation of the trustworthiness provision becomes possible for traditional measurements. It has been shown that the problem of trustworthiness related to soft measurements taking into consideration the growing interest in them, also becomes actual.

STUDY OF PERIODIC PROCESSES TAKING INTO ACCOUNT THE UNCERTAINTY STATE FACTOR OF THE ANALYZED PARAMETER

The paper describes the approach and methodology for the numerical assessment of the quality of periodic processes. With regard to their particular case the processes of selfoscillations, on the basis of the proposed approach, the problem of optimization of the control system is formulated and possible ways of solving it are shown. The practical implementation of the approach is possible when synthesizing a control system, as well as in the variant of automatic adaptive control during its operation. The proposed solutions are based on the use of methods of the theory of entropy potentials. The implementation of the proposed solutions will improve the quality of monitoring of periodic processes and the efficiency of regulation systems in selfoscillation modes.

ANALYSIS OF PARSING AND WEBSCRAPING TOOLS WITHIN THE FRAMEWORK DEVELOPMENT OF ARBITRATION INVESTMENT STRATEGY ON THE MARKET SPORT BETS

This article explores both popular and newly invented tools for extracting data from sites and converting them into a form suitable for analysis. The paper compares the Python libraries, the key criterion of the compared tools is their performance. The results will be grouped by sites, tools used and number of iterations, and then presented in graphical form. The scientific novelty of the research lies in the field of application of data extraction tools: we will receive and transform semistructured data from the websites of bookmakers and betting exchanges. The article also describes new tools that are currently not in great demand in the field of parsing and web scraping. As a result of the study, quantitative metrics were obtained for all the tools used and the libraries that were most suitable for the rapid extraction and processing of information in large quantities were selected.

USING SOCIAL MEDIA DATA FOR PROBABLE CALCULATIONS USING SCENARIOUS MODELING METHODS: RESULTS OF ELECTORAL CAMPAIGNS

The article presents the application of scenario modeling methods to assess the potential for scaling electoral support for political parties through digital communications (communities in social networks) based on data obtained from social networks. An analysis of communities in several social networks was carried out, various indicators were downloaded, reflecting the activity of both communities and users of such communities. Based on these data, various aggregates were calculated. Then a software package was developed that implements scenario modeling based on various identified indicators. The scenarios provide for the development of groups in social networks, depending on the activity of these groups. In this case, the activity is given by a random variable with a normal distribution. To test the developed algorithms, indicators of political communities in social networks were used.

ANALYSIS AND SYSTEMATIZATION OF THE FACTORS OF THE EXTERNAL ENVIRONMENT OF THE INVESTMENT PORTFOLIO BASED ON THE REGULARIZING BAYESIAN APPROACH

The article examines the factors and properties inherent in the external environment for companies – issuers of securities. These factors shape the environment and influence the behavior of the market as a whole, the situation in individual countries, regions, and industries. The systematization of factors is carried out, the causal relationships and the nature of the influence of individual factors on the behavior of securities quotes are indicated. A model of environmental factors based on a regularizing Bayesian approach is proposed.

POTENTIAL FEATURES OF THE MAXIMUM LIKELIHOOD METHOD

In this paper, the potential possibilities of the maximum likelihood method for solving statistical problems of radio engineering are analyzed. A condition is introduced for the correlation interval of a random vector of parameters of a set of signals, which expands the scope of the maximum likelihood method under conditions of a priori and parametric uncertainty. Proofs are given that the known methods of spectral, correlation analysis, and angular spectral analysis are special cases of the maximum likelihood method. The scope of their application for signal processing is limited. They determine the optimal estimates of the signal parameters only when the adopted implementation contains a single signal. The substantiation of the possibility of obtaining solutions to statistical problems of radio engineering in the field of nonorthogonality of signals, when spectral lines, correlation functions, radiation patterns partially coincide, is given. The issues of solving the problem of separate detection of a set of signals based on the proposed optimal receiver are discussed. The issues of separate estimation of signal parameters in the area of their nonorthogonality are discussed. The conditions of the maximum resolution of signals are analyzed. The possibility of creating maximum likelihood filters based on likelihood equations is discussed. It is shown that such filters allow separating signals in the area of their nonorthogonality. They are the basis for solving the problem of channel sealing in communication systems. They make it possible to develop communication systems with nonorthogonal carrier frequencies. A concrete example shows the possibility of exceeding the speed of information transmission in nonorthogonal communication systems in comparison with the maximum speed, which follows from the Shannon theorem. The paper presents the results of model calculations illustrating the provisions of the theory.

CREATION OF A SOFTWARE INTERFACE TO SUPPORT THE SOLUTION OF RESEARCH PROBLEMS IN THE FIELD OF SOFT COMPUTING USING THE PHYTON SCIKIT-FUZZY LIBRARY

The article discusses the issue of creating a software interface for developing fuzzy inference systems using the ScikitFuzzy library. With the help of the developed software module, a computational experiment was carried out to calculate the fuzzy output of the values of technological parameters for a fourcase evaporator unit, which is part of the evaporation shop of chlorine production.

DETERMINATION OF THE DEPENDENCES OF THE ANGLES IN THE JOINTS OF THE LOWER EXTREMITIES OF A HUMAN DURING GAIT

The study of human gait is of interest in determining its individual characteristics that affect the loads experienced by the joints of the lower extremities. To analyze the gait in the lateral projection, an experiment was carried out with placed markers on the subject's body to determine the interarticular angles in the hip, knee and ankle joints. The graphs of the dependences of these angles on time were built, corresponding to a certain phase of the step. It was found that the range of changes in the angle in the left ankle joint in the subject was significantly different from the dependence obtained for the right joint. Based on the subject's anamnesis, we associate these differences with the fact that there was an ankle joint injury that affected the condition of the ligamentous apparatus of the joint. In further work, it is proposed to consider a person's gait not only in the lateral, but also in the frontal (anteroposterior) plane to assess all movements in the joints, especially those associated with the abduction of the limbs and the human body from the vertical line on the specified plane.

MULTI-FACTOR ANALYSIS OF THE SCALES OF EVALUATING KNOWLEDGE BASED ON A FUZZY-PROBABILITY APPROACH

Teachers always face problems when assessing the competence level of test participants (whether students, undergraduates, etc.) in both moral and psychological fields as well as methodological. This raises important questions on the creation of professional systems, which have mathematical models set forth for their vital foundational path. The current working method in the creation of mathematical models for evaluating the competency of test participants is based on a fuzzy logic approach using both the explicit and implicit knowledge of highly qualified experts. One can see the realisation of this method in the example of Englishlanguage instruction. A comparison of two common assessment scales for knowledge of the tested has been carried out – the «fivepoint » standard adopted in Russia (or required, actually) and the converted sevenpoint ECTS (European Credit Transfer and Accumulation System) rating scale which is based on percentage. This study has shown that teachers cannot effectively assess the participants’ competency, neither on individual topics nor on the course as a whole, due to the narrowness of the «fivepoint » scale.