+7 (495) 987 43 74 ext. 3304
Join us -              
Рус   |   Eng


“Journal of Applied Informatics” is a peer-reviewed science journal with international representation of editorial board and authors, covering a significant part of Russian IT-area. The topics of the publications are connected to the aspects of theory and application of computer modeling and information technologies in various professional areas. The journal is indexed by Russian Science Citation Index on Web of Science platform.

In accordance with the decision of the Higher Attestation Commission of the Ministry of Education and Science of Russian Federation, journal is included in the «List of Leading Peer-Reviewed Scientific Journals and Publications authorized to publish main dissertation results»

A method for predicting bank customer churn based on an ensemble machine learning model

The results of research are presented, the purpose of which was to develop a method for predicting the outflow of clients of a commercial bank based on the use of machine learning models (including deep artificial neural networks) for processing client data, as well as the creation of software tools that implement this method. The object of the study is a commercial bank, and the subject of the study is its activities in the B2C segment, which includes commercial interaction between businesses and individuals. The relevance of the chosen area of research is determined by the increased activity of banks in the field of introducing digital services to reduce non-operating costs associated, in particular, with retaining clients, since the costs of attracting new ones are much higher than maintaining existing clients. The scientific novelty of the research results is the developed method for predicting the outflow of commercial bank clients, as well as the algorithm underlying the software that implements the proposed method. The proposed ensemble forecasting model is based on three classification algorithms: k-means, random forest and multilayer perceptron. To aggregate the outputs of individual models, it is proposed to use a learning tree of fuzzy inference systems of the Mamdani type. Training of the ensemble model is carried out in two stages: first, the listed three classifiers are trained, and then, based on the data obtained from their outputs, a tree of fuzzy inference systems is trained. The ensemble model in the proposed method implements a static version of the forecast, the results of which are used in a dynamic forecast performed in two versions – based on the recurrent least squares method and based on a convolutional neural network. Model experiments carried out on a synthetic dataset taken from the Kaggle website showed that the ensemble model has a higher quality of binary classification than each model individually.

Algorithm for identifying threats to information security in distributed multiservice networks of government bodies

The results of studies are presented, the purpose of which was to develop an algorithm for identifying information security threats in distributed multiservice networks that provide information interaction of regional government bodies, as well as their communication with the population of the region. The relevance of the research topic is due to a significant increase in various types of cyber attacks on the computer networks of public authorities and the need to increase the level of security of these networks by intellectualizing methods for combating information security threats. The algorithm is based on the use of machine learning methods to analyze incoming traffic in order to identify events that affect the state of information security of public authorities. The algorithm provides for input traffic preprocessing, as a result of which a set of images (signatures) obtained from Wasm binary files is formed, and then the image classifier is launched. It contains a sequential inclusion of deep neural networks – a convolutional neural network for signature classification and a recurrent network that processes the sequences obtained at the output of the convolutional network. Features of the formation of signatures in the proposed algorithm, as well as sequences at the input to the recurrent network, make it possible to obtain the resulting assessment of information security, taking into account the history of its current state. The output of the recurrent network is aggregated with the result of comparing the actual signatures with those available in the database. The aggregation is performed by the fuzzy inference system of the second type, using the implication according to the Mamdani algorithm, which generates the final assessment of information security threats. Software was developed that implements the proposed algorithm, experiments were carried out on a synthetic data set, which showed the efficiency of the algorithm, confirmed the feasibility of its further improvement.

Algorithms for composing efficient business models

Solving the problems of effective business management is associated with a variety of current goals facing the same and, by implication, requires the construction of appropriate models of efficient business. The article presents two problems of doing business which, apart from their common target being an improvement of business efficiency, have different current goals. The creation or development of any business involves the construction of a specific business plan for it, including a list of those areas of business development, the implementation of which will increase its efficiency. The first problem considered in the article is related to the phased implementation of all areas of efficiency improvement in order to ultimately obtain the greatest efficiency of their realization. The second one solves the problem of increasing efficiency by partially implementing efficiency improvement directions from the initial list, taking into account certain limitations, for example, in conditions of limited company resources. For the construction of models which would meet the problems set, an efficiency criterion is substantiated and proposed in the article, and Algorithms 1 and 2 are developed which made it possible to build the efficient business models which take into account the difference in its current goals. The authors have developed a multi-stage Algorithm 1 for the generation of individual sets of areas for improvement of efficiency to be used to solve the tasks at hand. Algorithm 2 implemented at each stage of Algorithm 1 has been developed by the authors by using the Pareto optimality method but supplemented by taking into account the features and objectives of the current tasks set for the business. The use of such algorithms has made it possible to build efficient business models enabling not only to obtain an economic effect inherent to each efficiency improvement area, but also to ensure additional growth thereof driven by the properties of the developed algorithms.

An approach to the design of a neural network for the formation of an individual trajectory of knowledge testing

The paper discusses the issues of implementing an adaptive testing system based on the use of artificial neural network (INS) modules, which should solve the problem of intelligent choice of the next question, forming an individual testing trajectory. The aim of the work is to increase the accuracy of the INS to form the level of complexity of the next test question for two types of architectures – direct propagation (FNN – Feedforward Neural Network) and recurrent with long-term short-term memory (LSTM – Long-Short Term Memory). The data affecting the quality of training are analyzed, the architectures of the input layer of the direct propagation INS are considered, which have significantly improved the quality of neural networks. To solve the problem of choosing the thematic block of the question, a hybrid module structure is proposed, including the INS itself and a software module for algorithmic processing of the results obtained from the INS. A study of the feasibility of using direct propagation ANNs in comparison with the LSTM architecture was carried out, the input parameters of the network were identified, various architectures and parameters of the ANN training were compared (algorithms for updating weights, loss functions, the number of training epochs, packet sizes). The substantiation of the choice of a direct distribution network in the structure of the hybrid module for selecting a thematic block is given. The above results were obtained using the Keras high-level library, which allows you to quickly start at the initial stages of research and get the first results. Traditionally, learning has taken place over a large number of eras.

Analysis and testing of neural network TCP/IP packet routing algorithms in private virtual tunnels

One of the most important components of the global Internet are traffic control and management systems. In order to achieve uninterrupted information and communication interaction, the organization of the process is constantly changing, covering not only individual subnets, but also p2p network architectures. The dominant areas for improving the network structure include 5G, IoT and SDN technologies, but their implementation in practice leaves the issue of ensuring the information security of networks built on their basis without a satisfactory solution. Current virtual tunnel deployment topologies and intelligent traffic distribution components provide only partial solutions, particularly in the form of access control based on user traffic and security through dedicated user certificates. The deployment of a tunnel is of particular importance in cases where it is necessary to ensure consistency and coordination of the work of complex socio-economic systems, an example of which is the information and communication exchange between participants in scientific and industrial clusters formed to implement projects for the creation of innovative products. However, existing solutions have disadvantages such as the need to purchase a license for full-featured access to the software product and specialized configuration of client-server authentication that provides secure access to a remote network route. The approach proposed by the authors, based on neural network distribution of traffic between clients of a private dedicated network, allows us to eliminate the noted shortcomings. Based on this principle, a multi-module system for intelligent packet routing was created and tested through unit testing. An analysis of the effectiveness of using a trained network address distribution model is presented in comparison with the use of a DHCP server based on the isc-dhcp-server package, distributed as the dhcpd service.

Building the mathematical model of the decision support system in the field of pricing for e-commerce

This work is devoted to the study of pricing issues for obtaining maximum profit when selling consumer goods at a constant purchase price. The said goods come in from either manufacturers or warehouses where the retail companies buy the goods in order to sell them directly to the consumers. The dependence of the selling rate per unit of time on the level of the added price in relation to the purchase price of the item is established by the means of sales price variation. The object of the research is the specific case of a linear approximation of said dependence, which is usually actualized in the event of either more elastic or less elastic demand for goods, when they are sold through Internet platforms. The proposed approach to determining prices of all the goods which are being sold for maximizing the total profit from the sales of all consumer goods or maximizing the total revenue throughout the whole period of sales time, based on the search of extremum points of the profit and revenue functions for each item of goods remains valid in the case of more complex approximations by quadratic and cubic functions of demand function. The type of the function of maximum value added revenue and the type of the function of maximum profit can be both found per unit of time depending on the variable level of the added price included into the sales price of the item. The type of maximum revenue function can be found per unit of time depending on the sales price of the item. The extremum points of the found functions are being determined. The theorems have been proved, that the extremum points which are being determined appear to be the maximum points of the researched functions for each item of goods, when the maximum profit or the maximum revenues are reached by selling goods to consumers. All common variables of said functions are found by summing up these functions among the multitude of goods on the interval of the whole sales time. The received data is used for the practical implementation of an effective sales strategy that ensures maximum profits for companies specializing in direct sales to consumers of the purchased goods. An applied methodicalэф approach to the sales of goods which ensures maximum profit from the sales in the field of elastic demand approximated by a linear function and under the condition of a constant purchase price for goods is proposed and theoretically substantiated.

Comparison of mathematical models of the dynamics of electrically charged gas suspensions for various concentrations of the dispersed component

The author: Tukmakov D.
This work is devoted to mathematical modeling of the dynamics of inhomogeneous electrically charged media. A dusty environment - solid particles suspended in a gas – was considered as an inhomogeneous medium. The mathematical model implemented a continuous approach to modeling the dynamics of inhomogeneous media. The complete hydrodynamic system of equations was solved for each component. The system of equations for the dynamics of each component included the equations of mass continuity, momentum components, and the energy conservation equation for the mixture component. Intercomponent interaction took into account momentum exchange and intercomponent heat transfer. The carrier medium was described as a viscous compressible heat-conducting gas. The flow was described as a flow with a two- dimensional geometry. The equations of the mathematical model were supplemented with initial and boundary conditions. The mathematical model took into account the wall viscosity in the channel. The system of equations of the mathematical model was integrated by McCormack's explicit finite-difference method. To obtain a monotonic grid function, a nonlinear scheme for correcting the numerical solution was used. The mathematical model was supplemented by the Poisson equation describing the electric field formed by charged dispersed particles. Poisson's equation was integrated by finite-difference methods on a gas-dynamic grid. Such a choice of the computational grid was necessary to calculate the concentration of particles required both for solving the electric field equation and for calculating the physical fields of the dynamics of inhomogeneous media. The reciprocal motion of a gas suspension caused by the movement of dispersed particles under the action of the Coulomb force was numerically investigated. The values of the surface and mass densities are determined, at which the models of the surface and mass densities of charges in the simulation of such a process are the same. It is revealed that the surface and mass models of charges are identical with respect to the volumetric content.

Computer program for modeling the technical quality indicators of inductors of linear asynchronous electric motors

The determination of indicators of the technical quality of electric motors, as well as their modeling, are in demand in the organization, final control of the technical quality of products, in the procedures for diagnosing and monitoring the technical condition of electrical equipment during operation. A large role in substantiating the criteria for the technical quality of electric motors and developing methods for their determination is given to computer modeling. Along with rotating electric motors, linear asynchronous electric motors (LAEM) are increasingly being used in technical systems. Criteria and test methods developed for rotating electrical machines are not adequate to the specifics of the LAEM. As applied to LAEM inductors, a method is proposed for assessing the technical quality indicators of three-phase inhomogeneous inductors with a non-uniform distribution of magnetic properties along the magnetic circuit. The method is based on the impulse effects of a voltage source on the LAEM inductor. A mathematical model, an algorithm and a program for modeling the technical quality indicators of LAEM inductors that implement the proposed method have been developed and presented. The program is written using its own interpreted programming language of the Maple software product and requires the user to enter the parameters of an electrical device. At the output, the user automatically receives an image of the Green’s matrix, oscillograms of phase currents based on the results of three experiments, and numerical values of quality indicators. The simulation results indicate that the proposed method can serve as a tool for the technical control of inductors, and also that the developed technical quality criteria, mathematical model, algorithm and program allow modeling and evaluating the technical qualities of LAEM inductors both during final tests and during monitoring the technical condition of the device during operation.