Modern Algorithms Research Articles

Active target classification using a shallow neural network with dimension reduction

Techniques for automatically identifying sonar targets using machine learning algorithms are being actively developed, and modern algorithms based on deep learning are often considered first. However, deep learning based algorithms require a lot of data to train a neural network, and if there is not enough data, overfitting easily occurs and the performance on real data can be degraded. In fact, in the case of active target detection where there is not enough data, it has been reported that using a shallow neural network after extracting appropriately designed features from the raw data showed better detection performance than applying deep learning directly to the raw data. With regard to this, we investigate the performance of a shallow neural network combined with dimensionality reduction techniques for active sonar target classification. In particular, several linear and nonlinear dimension reduction techniques are compared in terms of target classification performance, and the effects of the characteristics of background noise and the presence of reverberation on the target classification performance are discussed. [This work was supported by the research project PES4380 funded by KRISO.]

To the Problems of the National State Recognition System Improving

Introduction. A rapid increase in the number of objects that simultaneously take part in combat operations in the air requires improvement of systems for recognizing military objects both in terms of qualitative and quantitative indicators. This requires the development of appropriate algorithms for identifying new-generation "friend-foe" objects. Such algorithms can be based on various methods of information security, in particular symmetric and asymmetric cryptographic algorithms and other methods of cryptography. The purpose of the article is to survey modern systems of state recognition of objects (SSRO), identify their shortcomings and provide recommendations for their elimination. Results. The requirements for SSRO as a system for processing, transmitting, securing information and identifying objects based on cryptography and computer security methods are defined. Common and distinctive features for civil and military identification systems are highlighted. The advantages and disadvantages of the existing SSRO are shown. Recommendations are formulated to address the shortcomings of the existing SSRO. An example of a stand and a hardware and software basis for studying SSRO algorithms with backup channels is given. Conclusions. Eliminating the shortcomings of the SSRO and improving the level of its reliability will require the implementation of the following organizational and technical measures. 1. Replacing the current state recognition system with a more modern one, which will support more modern cryptographic algorithms and a larger number of recognition objects. Adding support for radio tag recognition (RF tags). 2. Support for recognition lines in various directions, including "ground – UAV", "plane – tank", "plane – UAV" and others. Adding support for the NATO standard – STANAG 4579, adopted in 2001, and others. 3. Using of broad-spectrum signals to reduce the probability of detection and interception, as well as a number of signal-code structures and a working frequency grid. Keywords: Friend-or-Foe, object identification, cryptography, backup channels.

Formation of the base and the automated system of the data integration about the industry

The expediency of forming the specialized database about industry and the system of their automated collection and usage were justified in this paper. This system allows to provide the information needs of the analysts and the scientists about the problems of industrial development. It provides the automated collection of the information for many countries of the world using any number of sources available as client-server resources on the Internet. The usage of modern data integration algorithms provides a diverse presentation of information, formats for its provision and frequency of updates. In addition, the advanced users of the system are provided with a wide range of options for creating search criteria and data acquisition format. The system is built as a client-server technology and is available for seamless integration with similar systems as a source and storage of information. In order to build a system, the analysis of existing in the world and popular in use statistical databases was carried out, their features, advantages and disadvantages were deаfined. It has been established that most of the existing statistical systems do not provide statistical data for Ukraine, and where they are, there is no detail data by type of economic activity and industry, and especially in industry. In addition, there are certain technical difficulties in working with data for users, the personalization of access is almost non-existent. The possibilities of the integrating existing databases with the systems that use information for modeling and forecasting both in query mode and especially in real time are rather limited. The structure of the statistical database about the development of industry has been formed, the indicators have been selected for its filling. The indicators were distributed by the sectors of the economy (the raw material production sector, the processing sector, the macrostatistics, other sectors). The sectors were disclosed by the key types of economic activity in accordance with the classification of economic activities in 2010 (CEA-2010), providing the predominant part of the creation of the added value of the economy of Ukraine. Each specific type of economic activity is disclosed by groups of industries and industries. In addition, the formation of the statistical database was carried out on the basis of assigning to each of the indicators other characteristics necessary for the automation: code, units of measurement, period, base (distribution of the indicators into the separate groups for the possibility of modeling), source (the statistical bases and sources from which the indicators and the data were collected). The modern relational database was used to store information, which allows solving optimization issues for working with the most powerful, but not yet large data, taking into account the features of wide data, allows horizontal and vertical scaling, including in the PostgreSQL open source database system. The methodology for the formation and the technology for filling the statistical database automatically from a large number of sources, the access mode to which is set by the system configuration parameters has been created. The technology has been developed for data migrations available in flexible formats, including text, in particular Excel. The implementation of the developed automated information system integration of industrial statistics allows you to select the necessary indicators for the analysis of economic processes in industry, use the statistical data collected in a single information space for 130 countries of the world (including Ukraine) for conducting scientific researches, building economic and mathematical models and making managerial

A Comparison Among Random Search Algorithms for PID Controller Optimization

Random search algorithms have been widely applied in practice to solve optimization problems. In the field of automatic control, there has been much research applying random search algorithms to optimize the parameter of Proportional – Integral - Derivative (PID) controller. The content of this research mainly stops at comparing the performance of the PID controller using the traditional tuning method with the optimization algorithm-based tuning method. Therefore, there is a need for research to compare the performance among random search algorithms for finding optimal parameters of the PID controller. In this paper, authors simulate five well-known and modern stochastic search algorithms including Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Whale Optimization Algorithm (WOA), Chemical Reaction Optimization (CRO), Teaching Learning Based Optimization (TLBO) to find optimal parameters of the PID controller for controlling the speed of the DC motor. The simulation results showed that the CRO algorithm is the most suitable to optimize PID controller parameters for the proposed problem.

A successful interdisciplinary collaboration for the management of pneumoperitoneum and pericardial effusion in a premature newborn

Objectives. Percutaneously central catheters are typically used in neonatal intensive care units (NICU) to provide parenteral nutrition and drug therapy administration. Their use is associated with specific complications: occlusions, infections, thrombosis, rupture, and migration, including rare life-threatening conditions such as pericardial effusion and cardiac tamponade. Material and methods. We present the case of a patient hospitalized in the Neonatology Department of “Sf. Andrei” County Clinical Emergency Hospital of Constanta, diagnosed with cardiac tamponade at 18 days of life. Outcomes. Premature male newborn, gestational age (GA) of 33 weeks, birth weight (BW) 2380 g, Apgar score 5/7, issuing from pregnancy without prenatal care, admitted in NICU with Continuous Positive Airway Pressure (CPAP) respiratory support. He required intensive care during the neonatal period and presented with sudden cardiac instability 18 days after the insertion of a peripherally inserted central catheter (PIC-Line). The echocardiography demonstrated severe pericardial effusion with evidence of cardiac tamponade. Successful urgent subxiphoid pericardiocentesis was performed. Totally 20 ml of fluid was collected, consistent with the composition of the hyperosmolar solution infused. Conclusions. Despite the fact that it is a rare complication, cardiac tamponade should be considered in any newborn with a PIC-Line who presents with sudden suggestive deterioration (cardiorespiratory instability, bradycardia, cyanosis and metabolic acidosis), which does not respond to resuscitation maneuvers, when lines are considered to be placed correctly. Implementing a modern algorithm, who is easy to learn and quick to perform, with the help of ultrasonography, represents the key of success for detecting urgent neonatal complications leading to sudden deterioration.

EVOLUTION OF FORMATION OF HOTEL AREAS IN THE CITIES OF WESTERN EUROPE XVIІІ ‒ EARLY XX CENTURIES

In the research, the urban development aspect of hotel areas in the context of severe urban changes that took place in the cities of Western Europe during the XVIIІ - early XX centuries were analyzed. It was investigated how due to external factors of influence new composite forms of the location of hotels in urban areas were formed and more modern algorithms of interaction between current urban planning realities and the formation of new, improved functional and planning proposals in the design, construction and operation of European hotels were formed in the second half of the XIX – early XX century. The scale of evolutionary changes has been constructed during this historical period and the main stages of the formation of the hotel industry in Europe have been studied. Four main compositional variants of urban areas were discovered, in which the hotels of that time were built: the frontal site, through the section, located on the corner and the hotel quarter. The prevailing forms of the location of hotels in these areas in the urban environment were analyzed, as well as the external factors of influence not only on the change of the urban construction concept were investigated, but also on the emergence of new architectural and planning decisions, which in their basic structure have survived to this day almost unchanged. Among the main features of the development of urban development, it should be noted the active development of the area of the former economic yard, which functionally began to be used directly by the hotel itself. This philosophy of increasing the room number and profitability of the institution has led to a gradual complication of the composite component and to the complete development of the hotel area with the selection of one or more internal atriums that were involved in separate public spaces and hotel areas.

Struct2Graph: a graph attention network for structure based predictions of protein–protein interactions

BackgroundDevelopment of new methods for analysis of protein–protein interactions (PPIs) at molecular and nanometer scales gives insights into intracellular signaling pathways and will improve understanding of protein functions, as well as other nanoscale structures of biological and abiological origins. Recent advances in computational tools, particularly the ones involving modern deep learning algorithms, have been shown to complement experimental approaches for describing and rationalizing PPIs. However, most of the existing works on PPI predictions use protein-sequence information, and thus have difficulties in accounting for the three-dimensional organization of the protein chains.ResultsIn this study, we address this problem and describe a PPI analysis based on a graph attention network, named Struct2Graph, for identifying PPIs directly from the structural data of folded protein globules. Our method is capable of predicting the PPI with an accuracy of 98.89% on the balanced set consisting of an equal number of positive and negative pairs. On the unbalanced set with the ratio of 1:10 between positive and negative pairs, Struct2Graph achieves a fivefold cross validation average accuracy of 99.42%. Moreover, Struct2Graph can potentially identify residues that likely contribute to the formation of the protein–protein complex. The identification of important residues is tested for two different interaction types: (a) Proteins with multiple ligands competing for the same binding area, (b) Dynamic protein–protein adhesion interaction. Struct2Graph identifies interacting residues with 30% sensitivity, 89% specificity, and 87% accuracy.ConclusionsIn this manuscript, we address the problem of prediction of PPIs using a first of its kind, 3D-structure-based graph attention network (code available at https://github.com/baranwa2/Struct2Graph). Furthermore, the novel mutual attention mechanism provides insights into likely interaction sites through its unsupervised knowledge selection process. This study demonstrates that a relatively low-dimensional feature embedding learned from graph structures of individual proteins outperforms other modern machine learning classifiers based on global protein features. In addition, through the analysis of single amino acid variations, the attention mechanism shows preference for disease-causing residue variations over benign polymorphisms, demonstrating that it is not limited to interface residues.

Modeling a Potential Plant Habitat Using Machine Learning Methods

The article is devoted to modeling the potential distribution area of the species Pulsatilla turczaninovii Kryl. et Serg. Plant ecological niche modeling is the process of building models using modern computer algorithms and bioclimatic data to predict the distribution range of plant species. The result of the simulation is a model that can be used to map the area of growth or residence of species, predict the range or analyze the impact of the environment on species.
 Data are required on both the presence of species and their absence in a particular territory to build effective models for predicting ecological niches of plants. View absence points (or background points) are not registered in databases, but can be generated using different approaches.
 This article describes the implementation of three approaches to selecting pseudo-absence points of species in an operationally divided territory and presents the result of modeling the potential distribution area of the species Pulsatilla turczaninovii Kryl. et Serg. using the random forest algorithm — the most popular way to build ensembles of decision trees. The software implementation of the model is carried out in the high-level Python programming language.

A new firefly algorithm with improved global exploration and convergence with application to engineering optimization

Firefly algorithm (FA) is a powerful and efficient meta-heuristic algorithm which has shown effective performance in the recent literature when applied to solving engineering optimization problems. FA imitates the flashing behavior of fireflies. FA generates solutions randomly and assumes them as fireflies. However, these algorithms may suffer from premature convergence and poor global exploration when used to optimize complex and high dimension engineering problems. Therefore, this study has proposed a novel FA, called firefly algorithm 1 to 3 (FA1→3), via different types of movements of fireflies in an attempt to improve the global exploration and convergence characteristics of FA. A comprehensive study has been carried out on the CEC2014 test functions to compare FA1→3 with the standard FA and several modern improved FA algorithms to validate its performance. The experimental results demonstrate that FA1→3 has achieved acceptable performance. In addition, it has been applied to six real-world engineering problems to show the optimization capability, robustness, and efficacy of FA1→3 in comparison with modern algorithms As per simulations, FA1→3 has provided suitable performance and higher accuracy than traditional and modified algorithms introduced in the last years. According to simulations, FA1→3 is significantly powerful and robust when dealing with various complex engineering problems and finds the design variables straightforwardly. Note that the source code of the proposed FA1→3 algorithm is publicly available at https://www.optim-app.com/projects/FA.

Machine Learning Model to Diagnose Diabetes Type 2 Based on Health Behavior

Diabetes, in 2016, was the 7th death-causing disease in the world. It was the direct cause of 1.6 million deaths. In 2019, the number of adults (20-79 years) that were living with diabetes was approximately 463 million and is expected to rise to 700 million in 2045. The early diagnosis of diabetes will help treat it and prevent its complications. The need for an easy and fast way to diagnose diabetes is crucial. In this study, we are proposing a method to diagnose diabetes with the help of machine learning algorithms and tools. The proposed method utilizes the power of machine learning to create a model that can predict diabetes based on the health behavior of the patient. The model uses the relationship between a healthy lifestyle and diabetes. Our goal is to build a reliable machine learning model to predict diabetes, which will help significantly in easing and speeding up the diagnosing procedure of diabetes. We used modern machine learning algorithms like XGBoost, LightGBM, CatBoost, and artificial neural networks, and the dataset was obtained from the National Health and Nutrition Examination Survey (NHANES). In our study, the XGBoost algorithm performed the best with a Cross-Validation (10-fold) score of 0.864, and an overall accuracy of 87.7% for the validation dataset and 84.96% for the test dataset.

How Do Spike Collisions Affect Spike Sorting Performance?

Recently, a new generation of devices have been developed to record neural activity simultaneously from hundreds of electrodes with a very high spatial density, both for in vitro and in vivo applications. While these advances enable to record from many more cells, they also challenge the already complicated process of spike sorting (i.e., extracting isolated single-neuron activity from extracellular signals). In this work, we used synthetic ground-truth recordings with controlled levels of correlations among neurons to quantitatively benchmark the performance of state-of-the-art spike sorters focusing specifically on spike collisions. Our results show that while modern template-matching-based algorithms are more accurate than density-based approaches, all methods, to some extent, failed to detect synchronous spike events of neurons with similar extracellular signals. Interestingly, the performance of the sorters is not largely affected by the spiking activity in the recordings, with respect to average firing rates and spike-train correlation levels. Since the performances of all modern spike sorting algorithms can be affected as function of the activity of the recorded neurons, scientific claims on correlations and synchrony should be carefully assessed based on the analysis provided in this paper.

Cardiac Biomarkers in 2022 – a Vital Tool for Emergency Care

Abstract The role of cardiac biomarkers in diagnosing acute myocardial infarction is undoubted. In the 2020 guidelines of the European Society of Cardiology, the measurement of cardiac peptides to gain prognostic information has a class IIa indication in all patients with ACS. In emergency care, ruling out a non-ST elevation myocardial infarction requires documentation of normal levels of cardiac biomarkers, which remain stable or have very small variations within several hours. This review aims to summarize the current knowledge and recent progresses in the field of cardiac biomarker discovery, from their routine use in emergency rooms to their prognostic roles in modern risk assessment tools. Integrated approaches combining cardiac troponin with other biomarkers of ventricular dysfunction or inflammation, or with modern cardiac imaging in emergency care are also presented, as well as the role of modern algorithms for serial troponin measurement in the modern management of emergency departments.

A fast GPU-accelerated Monte Carlo engine for calculation of MLC-collimated electron fields.

Although intensity-modulated radiation therapy and volumetric arc therapy have revolutionized photon external beam therapies, the technological advances associated with electron beam therapy have fallen behind. Modern linear accelerators contain technologies that would allow for more advanced forms of electron treatments, such as beam collimation, using the conventional photon multi-leaf collimator (MLC); however, no commercial solutions exist that calculate dose from such beam delivery modes. Additionally, for clinical adoption to occur, dose calculation times would need to be on par with that of modern dose calculation algorithms. This work developed a graphics processing unit (GPU)-accelerated Monte Carlo (MC) engine incorporating the Varian TrueBeam linac head geometry for a rapid calculation of electron beams collimated using the conventional photon MLC. A compute unified device architecture framework was created for the following: (1) transport of electrons and photons through the linac head geometry, considering multiple scattering, Bremsstrahlung, Møller, Compton, and pair production interactions; (2) electron and photon propagation through the CT geometry, considering all interactions plus the photoelectric effect; and (3) secondary particle cascades through the linac head and within the CT geometry. The linac head collimating geometry was modeled according to the specifications provided by the vendor, who also provided phase-space files. The MC was benchmarked against EGSnrc/DOSXYZnrc/GEANT by simulating individual interactions with simple geometries, pencil, and square beam dose calculations in various phantoms. MC-calculated dose distributions for MLC and jaw-collimated electron fields were compared to measurements in a water phantom and with radiochromic film. Pencil and square beam dose distributions are in good agreement with DOSXYZnrc. Angular and spatial distributions for multiple scattering and secondary particle production in thin slab geometries are in good agreement with EGSnrc and GEANT. Dose profiles for MLC and jaw-collimated 6-20-MeV electron beams showed an average absolute difference of 1.1 and 1.9mm for the FWHM and 80%-20% penumbra from measured profiles. Percent depth doses showed differences of <5% for as compared to measurement. The computation time on an NVIDIA Tesla V100 card was 2.5min to achieve a dose uncertainty of <1%, which is ∼300 times faster than published results in a similar geometry using a single-CPU core. The GPU-based MC can quickly calculate dose for electron fields collimated using the conventional photon MLC. The fast calculation times will allow for a rapid calculation of electron fields for mixed photon and electron particle therapy.

Kliuchovi aspekty transnatsionalnoi praktyky zapobihannia koruptsii ta perspektyvy yikh realizatsii v Ukraini

The relevance of scientific intelligence is that the existing problem of preventing corruption is not only a national problem, but also a global problem. The purpose of the article is to study the international experience in preventing corruption offenses in order to implement it in domestic law. The methodological basis of our scientific article is a system of methods and techniques of scientific knowledge, namely: systems analysis, comparative-implementation method, statistical method, retrospective method. The article provides a systematic review of international concepts for the prevention of corruption offenses. It is noted that in the current globalization processes that take place in modern society, the need to implement positive foreign experience in the system of current legislation of Ukraine is extremely important. A thorough analysis of a number of measures of the highly developed states aimed at preventing corruption offenses was also carried out. Positive experience of the states with the lowest level of corruption is positioned, indicating the ways to achieve such a result. The article analyses the anti-corruption strategies of such countries as Singapore, the Netherlands, Belgium, Israel, the United States of America, the Slovak Republic, Germany and Poland. Attention is drawn to the fact that in countries with low levels of corruption, repressive measures prevailing in anti-corruption models are combined with the comprehensive elimination of the determinants of corruption offenses. The article notes that the latest strategy to prevent corruption always requires the development of active cooperation between government agencies, law enforcement agencies and civil society in preventing and combating corruption offenses. At the same time, an important determinant of preventing corruption offenses is the growth of social consciousness. Emphasis is placed on the fact that corruption offenses are an extremely dangerous phenomenon, which, unfortunately, is common to all countries of the modern world. It is noted that a number of foreign countries have managed to create a modern and effective algorithm for preventing and combating corruption offenses. Taking into the consideration the European vector of our state, there is an urgent need to create and implement a new anti-corruption policy, taking into account the positive aspects of existing international experience.

A Survey on Analysis of Data Mining Algorithms for High Utility Itemsets

High-Utility-Itemset Mining (HUIM) is meant to detect extremely important trends by considering the purchasing quantity and product benefits of items. For static databases, most of the measurements are expected. In real time applications, such as the market basket review, company decision making and web administration organization results, large quantities of datasets are slowly evolving with new knowledge incorporated. The usual mining calculations cannot handle such complex databases and retrieve useful data. The essential task of data collection in a quantifiable sequence dataset is to determine entirely high utility sequences. The number of sequences found is always extremely high, though useful. This article studies the issue of the mining of repeated high utility sequence that meet item restrictions in order to identify patents that are more suited to the needs of a customer. Also, this article introduces high-value element set mining, examines modern algorithms, their extensions, implementations, and explores research opportunities.

The development of the motor and its speed regulation method

This paper first describes the types of motors and the corresponding speed regulation methods, and then introduces the motor speed regulation. PWM is regulated by PID, which regulate it from the beginning of the emergence of the parameters kp, ki, kd of the PID according to the empirical adjustment, to later development to adjust the parameters of the PID according to modern optimization algorithm. In this way, parameter control becomes accurate. Thus, the speed control of the motor is faster, which is of great significance for the application and development of the motor.

Completely quantum neural networks

Artificial neural networks are at the heart of modern deep learning algorithms. We describe how to embed and train a general neural network in a quantum annealer without introducing any classical element in training. To implement the network on a state-of-the-art quantum annealer, we develop three crucial ingredients: binary encoding the free parameters of the network, polynomial approximation of the activation function, and reduction of binary higher-order polynomials into quadratic ones. Together, these ideas allow encoding the loss function as an Ising model Hamiltonian. The quantum annealer then trains the network by finding the ground state. We implement this for an elementary network and illustrate the advantages of quantum training: its consistency in finding the global minimum of the loss function and the fact that the network training converges in a single annealing step, which leads to short training times while maintaining a high classification performance. Our approach opens a novel avenue for the quantum training of general machine learning models.

Modified Spatio-Temporal Matched Filtering for Brain Responses Classification

In this article, we apply the method of spatio-temporal filtering (STF) to electroencephalographic (EEG) data processing for brain responses classification. The method operates similarly to linear discriminant analysis (LDA) but contrary to most applied classifiers, it uses the whole recorded EEG signal as a source of information instead of the precisely selected brain responses, only. This way it avoids the limitations of LDA and improves the classification accuracy. We emphasize the significance of the STF learning phase. To preclude the negative influence of super–Gaussian artifacts on accomplishment of this phase, we apply the discrete cosine transform (DCT) based method for their rejection. Later, we estimate the noise covariance matrix using all data available, and we improve the STF template construction. The further modifications are related with the constructed filters operation and consist in the changes of the STF interpretation rules. Consequently, a new tool for evoked potentials (EPs) classification has been developed. Applied to the analysis of signals stored in a publicly available database, prepared for the assessment of modern algorithms aimed in EPs detection (in the frames of the 2019 IFMBE Scientific Challenge), it allowed to achieve the second best result, very close to the best one, and significantly better than the ones achieved by other contestants of the challenge.

A high-performance, hardware-based deep learning system for disease diagnosis.

Modern deep learning schemes have shown human-level performance in the area of medical science. However, the implementation of deep learning algorithms on dedicated hardware remains a challenging task because modern algorithms and neuronal activation functions are generally not hardware-friendly and require a lot of resources. Recently, researchers have come up with some hardware-friendly activation functions that can yield high throughput and high accuracy at the same time. In this context, we propose a hardware-based neural network that can predict the presence of cancer in humans with 98.23% accuracy. This is done by making use of cost-efficient, highly accurate activation functions, Sqish and LogSQNL. Due to its inherently parallel components, the system can classify a given sample in just one clock cycle, i.e., 15.75 nanoseconds. Though this system is dedicated to cancer diagnosis, it can predict the presence of many other diseases such as those of the heart. This is because the system is reconfigurable and can be programmed to classify any sample into one of two classes. The proposed hardware system requires about 983 slice registers, 2,655 slice look-up tables, and only 1.1 kilobits of on-chip memory. The system can predict about 63.5 million cancer samples in a second and can perform about 20 giga-operations per second. The proposed system is about 5–16 times cheaper and at least four times speedier than other dedicated hardware systems using neural networks for classification tasks.

International Experience in Preventing Corruption as a Vector for Creating a National Anti-Corruption Strategy in Ukraine

Given the European choice of Ukraine, there is an urgent need to create and implement a new anti-corruption policy, considering the positive aspects of the existing international experience. The relevance of the scientific investigation is conditioned by the fact that the existing problem of countering corruption is not only national but also global. The purpose of the study is to consider the international experience of preventing corruption torts to implement them in Ukrainian legislation. The methodological basis of the study is a system of methods and techniques of scientific cognition, namely: system analysis, comparative and implementation method, statistical method, and retrospective method. The study provides a systematic review of international concepts of prevention of corruption torts. It is noted that in the context of globalisation processes taking place in modern society, the need to implement positive foreign experience in the system of the current legislation of Ukraine becomes extremely relevant. A thorough analysis of a number of measures taken by leading states that are designed to prevent corruption offences was also carried out. The positive experience of states with the lowest level of corruption is positioned and ways to achieve such a result are indicated. The study analyses the anti-corruption strategies of such countries as Singapore, the Netherlands, Belgium, Israel, the United States, the Slovak Republic, Germany, and Poland. Attention is drawn to the fact that in countries with a low level of corruption, repressive measures are combined with a comprehensive elimination of the determinants of corruption offences in models of preventing corruption. The paper states that the latest strategy for preventing corruption requires the development of active cooperation between state bodies, law enforcement agencies, and civil society in order to prevent and counteract corruption torts. At the same time, an important determinant of preventing corruption torts is the growth of civil consciousness. It is emphasised that corruption offences are an extremely dangerous phenomenon inherent in all states of the modern world. It is revealed that a number of foreign countries have managed to create a modern and effective algorithm for preventing and countering corruption offences. The paper highlights the main determinants of corruption prevention that are approved by the international community, namely: normative regulation of the activities of civil servants, the establishment of a clear system of legal responsibility for violating the requirements of anti-corruption legislation, transparency in the professional activities of officials, the introduction of social programmes and educational campaigns on corruption topics. The provisions presented in the paper can become an effective basis for building a successful anti-corruption policy in Ukraine.