Search

Background. In European society, the high education, research and innovation are an essential component of socio-economic and cultural development. Especially they acquire in the field of high medical (pre- and postgraduate) education, since they significantly affect the health of society through training personnel in the health sector. The purpose was to develop the information model of the quality management system of the institution of high medical education on the basis of ISO 9001:2015 and its national analogue DSTU ISO 9001:2015. Results. The results of the system analysis of the structure and content of ISO 9001: 20015 and its national analogue DSTU ISO 9001: 20015 have allowed to establish that the basis of the construction of a quality management system, that meets the requirements of the standard, is based on the information exchange. Thus, one of the basic principles of quality management presented in the introduction to the standard, determines the principle of making managerial decisions based on factual data. This means that in order to effective quality management, the head of the institution of high medical education must have the relevant volume and content of information about the aims, objects and processes that form and maintain the system's effectiveness. Another guiding principle according to the standard - the process approach - provides, in particular, the continuous improvement of the quality of all processes (activities) in the management system based on the evaluation of data and information about their progress, interim and final results. The standard recommends keeping up-to-date documented information necessary for the functioning of the processes. In detail the detailed analysis of the structure and content of the international and national standards ISO 9001: 2015 and DSTU ISO 9001: 2015 requires the development of an information model of the quality management system for institutions of high medical education. Conclusions. Thus, the proposed information model of the quality management system of the institution of high medical education can serve as a benchmark for the managers of these institutions in the process of developing, implementing, supporting the functioning and evaluation of the quality management system with the aim of continuous improvement of the quality of high medical education at the postgraduate level.

One of the main elements of the lock’s mooring equipment is the floating bollard. Failure of this element of the mooring equipment always leads to great difficulties in organizing ship passes in the lock, in terms of the arrangement of vessels in the lock chamber and the forced withdrawal of locks from work to bring the floating bollards into working condition. Therefore, the analysis of the performance of this element, both at the design stage and during operation, taking into account various options for its loading under the action of changing external conditions, is an urgent task. The article presents the results of a study of one of the structures of a floating bollard of a ship’s lock, which is in real operation. The studies were carried out using the finite element method, in which 3D modeling of the device under consideration was carried out and its stress-strain state was analyzed under various loading options, taking into account the dynamics of mooring operations and weather conditions. A comparative analysis of the created finite element model of the device under consideration with its simplified model in the form of an elastic beam system with a uniform cross section is carried out. A graphic interpretation of the results of the analysis of the stress-strain state of the device under consideration is given, in which the most loaded structural elements are distinguished. The identified location of the places of the largest stress-strain state is proposed to be used as a theoretical basis for the deployment of sensors of the system for alerting the state of the loading of the floating bollard of the ship lock.

In this paper, an expert system model prototype based on Bayesian networks is proposed, which makes it possible to provide assistance to a doctor in the early diagnosis of a disease such as “pneumonia”. We designed a static Bayesian network with five key variables to obtain the probabilistic inference of the resulting node that determines the presence or absence of disease in a patient. We consulted with medical experts when selecting and quantifying input and output variables. When constructing a Bayesian model and conducting scenario analysis for a better prognosis of the diagnosis, we consulted with the attending physicians.

The event model for the processes of epidemic emergency localization is constructed. The obtained calculations show that the expected gain of time for the localization of the epidemic situation, using the event model, according to the forecast data is 36–40 hours. Разработана событийная модель развития эпидемии с учетом причинно-следственных связей между региональными компонентами. Решения таких задач включают в себя использование ситуационного управления и новых информационных технологий. Проведена оценка эффективности разработанной модели. Полученные расчеты показывают, что ожидаемый прирост времени, отведенного на локализацию чрезвычайной эпидемической ситуации, согласно прогнозируемым данным, составляет 36–40 часов. Розроблено подієву модель розвитку епідемічної надзвичайної ситуації з урахуванням причинно-наслідкових зв’язків між регіональними компонентами. Розв’язання таких задач включає в себе використання ситуаційного управління і нових інформаційних технологій. Проведено оцінку ефективності розробленої моделі. Отримані розрахунки показують, що очікуваний приріст часу на локалізацію надзвичайної епідемічної ситуації, згідно прогнозованим даним, становить 36–40 годин.

This paper presents the studies’ results on the probability-determined models development based on Bayesian networks to estimate the economic development measure of Ukraine. Considering that one of the difficulties in the Bayesian networks development is the exponential increase in the parameters amount in conditional probability tables (CPT), this study proposes a technique for applying Noisy-MAX nodes to model economic processes taking into account the time component. It is shown that if the proportion of enterprises that implement innovations is increased now by 31%, while the share of profits of these enterprises increases by only 2%, at the next time step the measure of manufacturability and innovation of Ukraine will rise by 81% and will tend to the maximum.

In the research, a dynamic BN (DBN) was designed to assess general trends in the level of regional competitiveness depending on economic detectors. This dynamic model is built on the basis of a trained, already verified static Bayesian network for assessing the country’s competitiveness. In contradistinction to an approach based on entrance from a numerical model as an entrance to a static network to foretell the meanings ​​of key detectors, the presented DBN uses the observed meanings ​​of key detectors to foretell future meanings ​​of these detectors taking into account the time component. Sensitivity analysis of BN is carried out, as well as a comparative analysis of “what-if” taking into account time steps. During the development of this DBN, it was found that the more evidence there is, the higher the accuracy of the designed network. The study identified baseline conditions, under which the competition detector at subsequent time steps will tend to increase. It is shown that to achieve the maximum level of competitiveness, we need to ensure the pursuit of maximum investment and innovation performance and improve the overall economic situation of the country.

This paper presents the results of studies on the development of probability-determined models based on Bayesian networks to assess the degree of economic development of Ukraine. Considering that one of the problems in the development of Bayesian networks is the exponential growth in the number of parameters in conditional probability tables (CPT), this study proposes a technique for using noisy-MAX nodes to model economic processes taking into account the time component. It is shown that if the proportion of enterprises that implement innovations is increased now by 31%, while the share of profits of these enterprises increases by only 2%, at the next time step the level of manufacturability and innovation of Ukraine will increase by 81% and will tend to the maximum.

In this paper, we propose a methodology for using dynamic Bayesian networks (DBN) in the tasks of assessing the success of an investment project. The methods of constructing DBN, their parametric learning, validation and scenario analysis of “What-if” are considered. A dynamic Bayesian model has been developed for scenario analysis and forecasting the success of an investment project. The model takes into account the time component and is designed in collaboration with expert economists in the selection and quantification of input and output variables. Now, using the dynamic Bayesian model, it is possible with a certain degree of probability to assess the degree of success of the capital investment, without incurring monetary and temporary losses. This will greatly facilitate the investment forecast for identifying profitable investment sources. This is the advantage of the proposed approach.

The article presents the results of research concerning development of inductive algorithm for hierarchical Bayesian clustering of gene expression of patients with two types of brain tumors and healthy individuals. The study carried out comparative studies of the clustering quality of inductive and classical methods of Bayesian hierarchical clustering algorithm (BHC). It is proposed to apply the moving average and FFT filtering methods for data dimension reducing. The basic principles of creating an inductive model of objective clustering are formed, the results of clustering are shown at various levels of data dimension reducing, the advantages of objective clustering BHC in comparison with the canonical BHC algorithm are determined.

The article considers the task of predicting the tertiary structure of a protein based on its primary sequence. The task of protein folding is solved by searching for conformation with minimal energy, if you look in terms of optimization problem. The work describes the use of a hybrid artificial immune system. In our work we try to combine clonal selection algorithms with differential and trigonometric differential evolution algorithms. We use special coding and decoding methods for individual individuals. The affinity function reduces the number of solutions with self-intersections (incorrect conformations) in the developed hybrid algorithms. We chose the hydrophobic-polar Dill model on a three-dimensional square lattice to test our algorithms. The study shows the advantages of the developed algorithms in comparison with other existing methods using experimental studies of test sequences.

A technique is proposed for grouping parts in a multinomenclature production based on a technical and economic analysis of the parameters of products and the manufacturability of parts manufacturing, which allows you to select a representative part of the group and calculate the required number and load factors of technological equipment. Testing the developed methodology at one of the machine-building enterprises showed that the deviation of the machine time for processing the “Plate” workpieces on metal-cutting machines from the estimated average is 12%.

In this paper, we propose a methodology for using static Bayesian networks (BN) in the tasks of predicting the development of the children’s disease “autism” aged 4 to 11 years. Methods of constructing the structure of a static BN, parametric learning, validation are considered, a sensitivity analysis and scenario analysis of “What-if’ are carried out. The model was designed in collaboration with pediatricians, as well as subject matter experts in the selection and quantification of input and output variables.

Time decrease in technological pre-production in multiproduct production is possible at the expense of the development of a method which takes into account current information of a production section.

This paper proposed a methodology for the use of static and dynamic Bayesian networks (BN) in the problems of localizing the distribution of narcotic substances. Methods for constructing the BN structure, their parametric training, validation, sensitivity analysis and “What-if” scenario analysis are considered. A model of dynamic Bayesian networks (DBN) for scenario analysis and prediction of the composition of a narcotic substance has been developed. The model was designed in collaboration with law enforcement officers, as well as forensic experts in the selection and quantification of input and output variables.

The Internet of Things (IoT) and Industrial Internet of Things (IIoT) are based on the wireless sensor networks (WSN), which include smart sensors and computer devices for monitoring and controlling processes in industry, agriculture, medicine, etc. In the networks, the sensed data are converted to the digital value of plants’ state, medical parameters of human, concentrations of matters in food products, such as juice, water, and so on. In the Institute of Cybernetics of National Academy of Sciences of Ukraine, there were designed and launched to the market the wireless smart sensor networks for agriculture, food industry and medicine application. There are ZigBee and Bluetooth compatible networks with different architectures: peer-to-peer (with mobile phone), star, and multihop tree-like networks. The main units of the designed sensor networks are patented wireless smart sensors for determination of the photosynthesis intensity in the plants, the concentrations of different matters in water or juice, and for measuring and monitoring medical parameters of human in real time. The features of design and application of different wireless networks of smart sensors and edge nodes are presented in the paper. The main contribution of this paper is the designing and fast-prototyping of WSN according to the requirements of application of WSN in beverage manufacturing with using two wireless protocols simultaneously for customer convenience.

The paper proposes a phased method of applying filtering algorithms, descriptor clustering. At the first stage, the features are reduced by sequential application of the moving average and FFT filtering algorithms and the reduction of the discretization step. At the second stage, for the selection of signs using the cluster analysis method X-means. At the final stage, regression models are constructed using the regulatory regression algorithms L1, L2, and Leastsquares. The resulting models are highly accurate, robust and adequate. In general, the work proposed a new method for predicting the binding affinity of peptides in order to find the numerical values of peptide bonds.

In this paper, we propose a methodology for using static Bayesian networks (BN) to simulate the processes of educational institutions financing and their impact on Gross Domestic Product (GDP). The methods for constructing the BN structure, their parametric learning, validation, sensitivity analysis and scenario analysis «What-if» are considered.

The article is devoted to the information technology of using dynamic Bayesian networks (DBN) for assessing and modeling the quality of law enforcement agencies' work. A DBN model has been developed for analyzing and predicting the criminal situation in the region. The model was designed in collaboration with law enforcement officers, as well as forensic experts in the selection and quantification of input and output variables. The model was validated on a data set for the period from 2009 to 2019.

Modern precision and digital agriculture for efficiency increasing requires development and implementation of new information and communication technologies, including wireless ones. The authors developed wireless sensor network (WSN) for express-diagnostics of agricultural plants' state, created the appropriate hardware, software and methodological support. The developed network now is implemented and under field tests both in open agricultural lands and in smart greenhouses. The main contribution of this paper is the design and fast-prototyping of WSN, taking into account the requirements of application task concerning field usage, long time of autonomous operation and measuring environmental parameters. That caused modeling of developed network to estimate its parameters before practical realization.

This paper proposed a methodology for the use of static Bayesian networks (BN) in the problems of localizing the distribution of narcotic substances. Methods for constructing the BN structure, their parametric training, validation, sensitivity analysis and “What-if’ scenario analysis are considered.

Unit self-diagnosis is considered at system level. As distinct from system level self-diagnosis based on units mutual tests, we have researched the method based on the tests which a unit performs on other system units. Taking into account the obtained test results, a unit evaluates its own state. In our research, we have considered different faulty assumptions and testing procedures. Diagnosis model was developed and analyzed. Computer simulation is performed by using the web application developed for this research. Results of simulation were analyzed and assessed. Some recommendations were made for achieving better diagnosis results.

In this paper, for solving the problem of forecasting non-stationary time series, hybrid learning methods for GMDH-neural networks are proposed. Training methods combine artificial immune systems with members of the evolutionary algorithm family, in particular, gene expression programming systems. The following hybrid computational methods for the synthesis and training of GMDH-neural networks have been developed: a method in which candidate models are represented as gene expression, and training is performed by clonal selection; the method of two-phase structural-parametric synthesis, in which the structural component is formed by programming the expression of genes, and the parameterization is performed by clonal selection; a method based on cooperative-competitive processes of interaction of the elements of the immune system, in which the structure and parameters of the GMDH-neural network are represented by the entire population element-wise. Comparative experimental studies of the quality of the proposed computational methods for solving forecasting problems were carried out.

The paper deals with the problem of protein tertiary structure prediction based on its primary sequence. From the point of view of the optimization problem, the problem of protein folding is reduced to the search for confirmation with minimal energy. To solve this problem, a hybrid artificial immune system has been proposed in the form of a combination of clonal selection and differential evolution algorithms. The developed hybrid algorithm uses special methods of encoding and decoding individuals, as well as an affinity function, which allows reducing the number of incorrect conformations (solutions with self-intersections). To test the algorithm, Dill’s hydrophobic-polar model on a two-dimensional square lattice was chosen. Experimental studies were conducted on test sequences, which showed the advantages of the developed algorithm over other existing methods.

The paper proposes a methodology for using static Bayesian networks (BN) to assess the factors influencing the assessment of the quality of the national law enforcement agencies work. The methods for constructing the BN structure, their parametric learning, validation, sensitivity analysis and scenario analysis «What-if» are considered.

System level diagnosis is an abstraction of high level and, thus, its practical implementation to particular cases of complex systems is the task which requires additional investigations, both theoretical and modeling. Mostly, diagnosis at system level intends to identify only permanently faulty units. In the paper, we consider the case when both permanently and intermittently faulty units can occur in the system. Identification of intermittently faulty units has some specifics which we have considered in this paper. We also suggest the method which allows distinguishing among different types of intermittent faults. Diagnosis procedure was suggested for each type of intermittent faults.

Digital agriculture requires development and implementation of new information and communication technologies including wireless ones for increasing efficiency. The authors developed wireless senso...

The article deals with the strength calculation issues of products by example of rammer under short-term loads in hammer drill. Based on finite element analysis’ results of the rammer geometrical model, it is proposed to select a material with a strength limit significantly exceeding the allowable residual stresses inside the rammer. The harmonic analysis allows us to estimate the rammer’s strength. The dynamic coefficient may be calculated by the use of Vibran 2.0 device (amplitude and signal frequency) and by calculation results of the maximum displacement. The FreeCAD software package was used for this. It allows to create internal scripts to automatically search for the optimal design parameters. A method of determining mechanism’s static load and its dynamic coefficient under periodic impact loads has been developed. Static loads calculations and harmonic analysis according to the developed method allow to determine structural strength of the product, to select the rammer’s material and to find the dynamic coefficient (which takes into account the stroke loads in calculations). The pro-posed calculation method may be used not only to solve the describing problem, but also for a whole class of problems related to stroke loading, where such parameters as application time of force or duration time of force’s impulse are important.

The paper deals with the problem of system level self-diagnosis. Usually it is assumed that a system unit is tested when it is idle and, thus. it can participate in a test. It means that a system unit cannot be tested continuously. In view of this, there is a probability of system unit fault omission. In the paper, we show how the tests of the system units are organized and proposed the method of evaluating the credibility of the results of system checking.

Binding of peptides plays an important role in the molecular biology. The binding process of peptides can activate the cytotoxic T-cells in the immune system. One of the most difficult aspects of peptide binding process is the prediction of the affinity of protein binding to the peptide. In this paper, we propose a new information technology for the quantitative determination of the affinity of peptide-MHC binding. In the paper a step-by-step method of applying descriptor clustering algorithms has been proposed. At the first stage we use the DBSCAN density algorithm, and the descriptors representing the ghost component with respect to the dependent variable are deleted. At the second stage, a robust set of prognostic features is identified using k-means algorithm. At the final stage, polynomial models are constructed with the use of iterative multilayer algorithm MIA GMDH. The obtained models have high accuracy, robustness, and adequacy.

The information technology of the operative formation of the method of the relevant alternatives of decision-making in human-machine systems critical application is offered. It is suggested to use the naive Bayesian classifier for the operative formation and evaluation of the relevant alternatives of acquisition in a situation of uncertain, inaccurate and inconsistent information in real time.

The paper presents the research concerning comparison analysis of biclustering algorithms effectiveness with the use of artificial data and gene expression profiles. Internal biclustering quality criterion is proposed as the result of the simulation. The change of this criterion has high correlation with Jaccard index, which was calculated for perfect and obtained biclustering. The technology of bicluster analysis based on “ensemble” method was proposed as the structural block-chart of step-by-step information processing to determine the optimal biclustering level using internal biclustering quality criterion.

System level diagnosis is an abstraction of high level and, thus, its practical implementation to particular cases of complex systems is the task which requires additional investigations, both theoretical and modeling. Traditionally, system self-diagnosis is used for detecting of permanently faulty nodes. In the paper, we consider the problems of intermittent fault detection and suggest diagnosis procedures which allow distinguishing between different types of intermittent faults. For each type of intermittent faults the diagnosis procedure was developed.

The paper presents the information technology of gene expression profiles processing in order to reconstruct gene regulatory networks. The information technology is presented as a structural block-chart, which contains all stages of studied data processing. DNA microchips of patients, which were studied on different types of diseases, were used as experimental data. The relative criteria of validation for all reconstructed networks were calculated during simulation process. The obtained results show high efficiency of the proposed technology. High values of the validation criteria indicate a high level of the obtained gene networks objectivity.

Two variations of the network are described i.e. Larsen model and Tsukamoto model. In each case an artificial immune network is proposed as a means of building fuzzy network structure and performing parameters inference.

In the article the features of wireless smart biosensor for sensor networks in ecological monitoring are considered. The developed wireless smart sensors manufactured on certified contract manufacture, is described. The results of wireless sensors testing are given.

This article gives an overview on the techniques needed to implement the discrete Fourier transform (DFT) efficiently on current multicore systems. The focus is on Intel-compatible multicores, but we also discuss the IBM Cell and, briefly, graphics processing units (GPUs). The performance optimization is broken down into three key challenges: parallelization, vectorization, and memory hierarchy optimization. In each case, we use the Kronecker product formalism to formally derive the necessary algorithmic transformations based on a few hardware parameters. Further code-level optimizations are discussed. The rigorous nature of this framework enables the complete automation of the implementation task as shown by the program generator Spiral. Finally, we show and analyze DFT benchmarks of the fastest libraries available for the considered platforms.

In the article the applications of wireless technologies in agriculture, ecological monitoring and defense are considered. The common approach of realization of wireless data acquisition systems for monitoring of biological object states in different areas of human activity was developed.

Fast changing, increasingly complex, and diverse computing platforms pose central problems in scientific computing: How to achieve, with reasonable effort, portable optimal performance? We present SPIRAL, which considers this problem for the performance-critical domain of linear digital signal processing (DSP) transforms. For a specified transform, SPIRAL automatically generates high-performance code that is tuned to the given platform. SPIRAL formulates the tuning as an optimization problem and exploits the domain-specific mathematical structure of transform algorithms to implement a feedback-driven optimizer. Similar to a human expert, for a specified transform, SPIRAL "intelligently" generates and explores algorithmic and implementation choices to find the best match to the computer's microarchitecture. The "intelligence" is provided by search and learning techniques that exploit the structure of the algorithm and implementation space to guide the exploration and optimization. SPIRAL generates high-performance code for a broad set of DSP transforms, including the discrete Fourier transform, other trigonometric transforms, filter transforms, and discrete wavelet transforms. Experimental results show that the code generated by SPIRAL competes with, and sometimes outperforms, the best available human tuned transform library code.

It the article it is described wireless devices and systems for application in precision farming, developed in V.M.Glushkov Institute of Cybernetics. The chlorophyll fluorescence induction is measured to fulfill express-diagnostics of plane state.

We present the automatic synthesis of the HPC Challenge’s Global FFT, a large 1D FFT across a whole supercomputer system. We extend the Spiral system to synthesize specialized single-node FFT libraries that combine a data layout transformation with the actual on-node FFT computation to improve the network performance through enabling all-to-all collectives. We run our optimized Global FFT benchmark on up to 128k cores (32 racks) of ANL’s BlueGene/P “Intrepid” and achieved 6.4 Tflop/s, outperforming ANL’s 2008 HPC Challenge Class I Global FFT run (5 Tflop/s). Our code was part of IBM’s winning 2010 HPC Challenge Class II submission. Further, we show first single-thread results on BlueGene/Q.

This paper presents an extension of the Spiral system to automatically generate and optimize FFT algorithms for the discrete Fourier transform over finite fields. The generated code is intended to support modular algorithms for multivariate polynomial computations in the modpn library used by Maple. The resulting code provides an order of magnitude speedup over the original implementations in the modpn library, and the Spiral system provides the ability to automatically tune the FFT code to different computing platforms.

Automatic tuning has emerged as a solution to provide high-performance libraries for fast changing, increasingly complex computer architectures. We distinguish offline adaptation (e.g., in ATLAS) that is performed during installation without the full problem description from online adaptation (e.g., in FFTW) that is performed at runtime. Offline adaptive libraries are simpler to use, but, unfortunately, writing the adaptation heuristics that power them is a daunting task. The overhead of online adaptive libraries, on the other hand, makes them unsuitable for a number of applications. In this paper, we propose to automatically generate heuristics in the form of decision trees using a statistical classifier, effectively converting an online adaptive library into an offline one. As testbed we use Spiral-generated adaptive transform libraries for current multicores with vector extensions. We show that replacing the online search with generated decision trees maintains a performance competitive with vendor libraries while allowing for a simpler interface and reduced computation overhead.