Objectives. The problem of constructing a new class of physically unclonable functions of the arbiter type (APUF) that combines the advantages of both classical and balanced APUF is solved. The relevance of such a study is associated with the active development of physical cryptography. The following goals are pursued in the work: research and analysis of classical APUF, construction of a new mathematical model of APUF and development of a new basic element of APUF.

Methods. The methods of synthesis and analysis of digital devices are used, including those based on programmable logic integrated circuits, the basics of Boolean algebra and circuitry.

Results. It has been established that classical APUF uses a standard basic element that performs three functions, namely, the function of generating two random variables Generate, the function of choosing a pair of paths Select and the function of switching paths Switch, which are specified by one bit of the challenge. It is shown that the joint use of these functions, on the one hand, makes it possible to achieve high characteristics of the APUF, and on the other hand, leads to the formation of an asymmetric behavior of the APUF. In order to analyze the main characteristics of APUF and their ideal behavior, a new mathematical model of APUF was considered, similar to the model of random coin toss. To implement APUF functioning according to the proposed model, a new basic element was developed. It is shown that the use of the proposed basic element allows to build symmetrical physically unclonable functions (C_APUF), which differ from the classical APUF in that the Generate, Select and Switch functions of the basic element are performed by their independent components and are specified by different bits of challenge.

Conclusion. The proposed approach to the construction of symmetrical physically unclonable functions, based on the implementation of the Generate, Select and Switch functions by various components of the base element, has shown its efficiency and promise. The effect of improving the characteristics of similar C_APUF has been experimentally confirmed, and, first of all, a noticeable improvement in their probabilistic properties expressed in equal probability of responses. It seems promising to further develop the ideas of building C_APUF, experimental study of their characteristics, as well as analysis of resistance to various types of attacks, including using machine learning.

Objectives. The problem of choosing the best methods and programs for circuit implementation as part of digital ASIC (Application-Specific Integrated Circuit) sparse systems of disjunctive normal forms (DNF) of completely defined Boolean functions is considered. For matrix forms of sparse DNF systems, the ternary matrix specifying elementary conjunctions contains a large proportion of undefined values corresponding to missing literals of Boolean input variables, and the Boolean matrix specifying the occurrences of conjunctions in DNF functions contains a large proportion of zero values.

Methods. It is proposed to investigate various methods of technologically independent logical optimization performed at the first stage of logical synthesis: joint minimization of systems of functions in the DNF class, separate and joint minimization in classes of multilevel representations in the form of Boolean networks and BDD representations using mutually inverse cofactors, as well as the division of a system of functions into subsystems with a limited number of input variables and the method of block cover of DNF systems, focused on minimizing the total area of the blocks forming the cover.

Results. When implementing sparse DNF systems of Boolean functions in ASIC, along with traditional methods of joint minimization of systems of functions in the DNF class, methods for optimizing multilevel representations of Boolean function systems based on Shannon expansions can be used for technologically independent optimization, while separate minimization and joint minimization of the entire system as a whole turn out to be less effective compared with block partitions and coatings of the DNF system and subsequent minimization of multilevel representations. Schemes obtained as a result of synthesis using minimized representations of Boolean networks often have a smaller area than schemes obtained using minimized BDD representations.

Conclusion. For the design of digital ASIC, the effectiveness of combined approach is shown, when initially the block coverage programs of the DNF system is used, followed by the use of programs to minimize multilevel block representations in the form of Boolean networks minimized based on Shannon expansion.

Objectives. The problem of developing an algorithm for estimating the absolute total electron content of the ionosphere from dual-frequency phase and range satellite measurements for a single receiving station of global navigation satellite systems is being solved.

Methods. To obtain an estimate the phase measurement data are corrected using digital signal processing methods, well known total electron content formulas for phase and range measurements are applied and combined, and also the differential code bias of the receiving station is estimated using the least squares method.

Results. It is shown that the total electron content calculated from phase measurements provides high accuracy, but up to an unknown constant, but the content calculated from range measurements allows one to obtain the absolute value, but with a large noise component and differential code bias of a satellite and receiver equipment. An algorithm for estimating the absolute total electron content of the ionosphere has been developed, its description and diagram are given. The algorithm was used to estimate the total electronic content within six months of observations, and the average error of the resulting estimate was calculated.

Conclusion. The developed algorithm can be used to estimate the absolute total electron content of the ionosphere for a single receiving station of global navigation satellite systems. In contrast to theoretically known formulas for phase and range measurements, this article contains information about adjusting phase measurements and estimating the differential code delay of receiving station. Further research may be related to the adaptive selection of parameters and testing of the algorithm for working with nanosatellites of the CubeSat format.

Objectives. The goals are to develop a nonlinear risk model and examine its prediction applicability for clinical use.

Methods. Methods of survival analysis and regression statistical models were used.

Results. A practical approach to assessing nonlinear risks of adverse events using the example of gastric cancer treatment is proposed. A model for predicting the metachronous peritoneal dissemination in patients undergoing radical surgery for gastric cancer was proposed and studied. Assessment of risks for various periods of observation was performed, and the clinical suitability of developed approach was assessed.

Conclusion. In clinical oncological practice, not only timely treatment plays an important role, but also the prevention of adverse outcomes after treatment. Individualization of patient monitoring after treatment reduces the risks of fatal outcomes and the costs of additional research and treatment in the event of cancer progression. Based on the results of this study, we propose solutions that should lead to more effective and high-quality treatment tactics and follow-up after treatment for gastric cancer, also to the selection of optimal approaches and to obtaining clinically favorable outcomes of the disease. The proposed risk prediction method will ultimately lead to individualized patient management based on the results of personal data.

Objectives. The purpose of the study is to construct and study the use of a feed-forward neural network to solve the problem of loan classification, as well as to conduct a comparative analysis of the neural networkbased approach with the existing approach based on logistic regression.

Methods. Based on a feed-forward neural network using historical data on loans issued, the following metrics are calculated: cost function, Accuracy, Precision, Recall, and measure, calculated on Precision and Recall values. Polynomial parameters and the principal component method are used to determine the optimal set of input data for the studied neural network.

Results. The impact of data normalization on the final result was analyzed, the influence of the number of units in the hidden layer on the outcome was evaluated using a two-stage method and the Monte Carlo method, the effect of balanced data use was determined, the optimal threshold value for output layer of the neural network under investigation was calculated, the optimal activation function for the hidden layer units was found, the effect of increasing input indicators through missing values imputation and the use of polynomials of varying degrees was studied and the redundancy in the existing set of input indicators was analyzed.

Conclusion. Based on the results of the research, we can conclude that the use of a direct distribution network to solve problems of loan classification is appropriate. Compared to logistic regression, implementing a solution using a feed-forward neural network requires more time and computing resources. However, the obtained most important values of Accuracy and measure are higher than those calculated using logistic regression [1].

Objectives. Currently, the main source of information is the Internet. The huge amount of information available on the Internet makes it urgent to comprehensively analyze data from open Internet sources.

The goal of this work is to create a multi-purpose, modifiable cluster for in-depth analysis of data from Internet sources, the main objectives of which are to identify the most important publications in a certain subject area, thematic analysis of these publications, identifying the leader of a scientific direction and determining trends in the development of areas and interaction of groups of people.

Methods. To solve this problem, a methodology was developed for constructing a multi-purpose cluster using technologies for quickly constructing a thematic graph database, a knowledge graph, methods and models of machine learning for in-depth analysis of data.

Results. A system for comprehensive analysis of data from thematic sites ISKAD IS has been developed, a methodology for quickly constructing a thematic graph database and a comprehensive technology for in-depth analysis of data from Internet sources and analysis of data from the most important well-known world sites have been tested.

Conclusion. An IT environment has been created for the rapid construction of thematic graph databases. The results of using the technology for quickly constructing graph databases are shown using examples of the work of ISKAD IS.