Objectives. The problem of developing a generative hetero-encoder model for computer-aided design of potential inhibitors of Bcr-Abl tyrosine kinase, an enzyme whose activity is the pathophysiological cause of chronic myeloid leukemia, is being solved.

Methods. A generative hetero-encoder model was designed based on the recurrent and fully connected neural networks of direct propagation. Training and testing of this model were carried out on a set of chemical compounds containing 2-arylaminopyrimidine, which is present as the main pharmacophore in the structures of many small-molecule inhibitors of protein kinases.

Results. The developed neural network was tested in the process of generating a wide range of new molecules and subsequent analysis of their chemical affinity for Bcr-Abl tyrosine kinase using molecular docking methods.

Conclusion. It is shown that the developed neural network is a promising mathematical model for de novo design of small molecules which are potentially active against Bcr-Abl tyrosine kinase and can be used to develop effective broad-spectrum anticancer drugs.

Objectives. For a method for estimating the total electron content in the ionosphere based on the retransmission of the L1 GPS navigation signal by a repeater nanosatellite (SR) to the frequencies of 150/400 MHz allocated for geophysical research and their reception at a ground-based receiving point (RP), it is necessary to develop the algorithms for coherent accumulation of received relayed signals and measurement of the difference between their delays at observation intervals up to a few seconds.

Methods. The proposed algorithms provide phase demodulation of the received signals at each of the relay frequencies in accordance with the dynamics of the mutual spatial movement of the navigation satellite (NS), SR and RP; multiplying the result by the estimate of the navigation message combined time delay, generated by the receiver of the direct navigation signal on the NS-RP route, intra-period processing over the entire duration of the observation in a filter matched with the signal of the navigation satellite, and inter-period coherent accumulation of the results of intra-period processing at a single-valued range interval. Coherent accumulation, taking into account a random uncontrolled shift in the frequency of the retransmitted signal, is implemented by discrete Fourier transform of the vectors formed for each resolution element in the delay time from inter-period readings of the results of intra-period processing, taking into account the range migration during the mutual movement of the NS, SR and RP.

Results. It is shown that by the output signal of the coherent accumulator makes it is possible to detect retransmitted signals at each of the retransmission frequencies, to measure accurately the difference in delay times, and estimate the total electron content on the SR-RP path. The results of modeling are presented, confirming the efficiency of the proposed algorithms in estimating the total electron content on the SR-RP route.

Conclusion. An algorithm for coherent accumulation of received retransmitted signals and measurement of the difference between their delays is developed, and its simulation is performed. The algorithm can be used for estimating TEC based on the retransmission of signals from GPS.

Objectives. Development of new approach for recognizing the fabric composition of clothing in e-commerce images by using generative adversarial network(GAN) to generate synthetic images of clothing with known fabric composition, to be used to train the CNN to classify the fabric composition of real clothing images. Instead of a classic clothing image, a copy is generated with the material zoomed to fibers and fabric structure.

Methods. The main methods to recognize the fabric composition of the clothing image in the e-commerce are the creation and annotation of a dataset for the neural network training, synthesis of the fabric of clothing, the choice of architecture and its modification, validation and testing, and interpretation of the results.

Results. Experimental results with the constructed method show that it is effective for accurately recognizing the fabric composition of e-commerce clothing to be used to improve search and browsing on websites.

Conclusion. In the course of the experiment, using a generative adversarial network, a data set of e-commerce products was synthesized and annotated, neural networks were built to recognize the composition of the fabric of clothing items. The results of the study showed that the new approach for recognizing the fabric of clothing provides higher accuracy in comparison with already known methods, in addition, the use of the attention model also gives good results to improve the metrics.

Objectives. The problem of investigating a fork-join queuing system is considered. It is required to build the process of the system functioning, to find the condition for the existence of a stationary distribution, and propose algorithms for calculating the stationary distribution and the main stationary performance characteristics. The special interest of the study is to obtain the lower and upper bounds of the mean sojourn time of a customer in the system.

Methods. Methods of probability theory, queuing theory and matrix theory are used.

Results. The functioning of the system is described in terms of a multidimensional Markov chain. A constructive condition for the existence of a stationary distribution is found, and algorithms for calculating the stationary distribution and stationary performance characteristics of the system are proposed. Analytical expressions are obtained for the lower and upper bounds of the mean sojourn time of customers in the system.

Conclusion. The functioning of the fork-join queuing system with a stationary Poisson flow has been studied. Each of the arriving customers forks into two tasks that go to two subsystems, each of which consists of a server and a buffer. We assume that the buffer to one of the servers is unlimited, and to the second server has a finite volume. Service times have, generally speaking, different phase distributions (PH-Phase type distributions). For this system, a condition for the existence of a stationary distribution is obtained, algorithms for calculating the stationary distribution and a number of stationary performance measures of the system are proposed. Analytical expressions for the lower and upper bounds of the mean sojourn time of a customer in the system from the moment it enters the system to the moment of synchronization, which is a critical performance indicator of the fork-join queues, are obtained. The results of the study can be used for modeling various computer and communication systems, in particular, systems that perform parallel computing, customer processing in distributed databases, and parallel disk access.

Objectives. Analytical solution of the boundary value problem of electrostatics for modeling the electrostatic field of a charged ring located inside a grounded infinite circular cylinder in the presence of a perfectly conducting torus is considered. The field source is a thin charged ring located on a plane perpendicular to the axis of the cylindrical screen.

Methods. To solve the problem, the method of addition theorems is used. The potential of the initial electrostatic field is presented in the form of spherical harmonic functions and in the form of a superposition of cylindrical and toroidal harmonic functions, using addition theorems relating spherical, cylindrical and toroidal harmonic functions. The secondary potential of the electrostatic field is also represented as a superposition of cylindrical and toroidal harmonic functions.

Results. The solution of the formulated boundary problem is reduced to the solution of an infinite system of linear algebraic equations of the second kind with respect to the coefficients included in the representation of the secondary field. The influence of some parameters of the problem on the value of the electrostatic potential inside a grounded cylindrical shield in the presence of a toroidal inclusion is numerically studied. The calculation results are presented in the form of graphs.

Conclusion. The proposed technique and the developed software can find practical application in the development and design of screens in various fields of technology.

Objectives. Models and tools for designing adaptive user interfaces for intelligent systems are being developed. The relevance is determined by the need to reduce overhead costs and development time for user interfaces and to provide their adaptation to the specific characteristics of the user of the intelligent system.

Methods. Existing approaches to designing user interfaces are being analyzed. A semantic model of an adaptive user interface for intelligent systems is proposed, implemented using a basic universal language for representing knowledge based on set theory and graph theory.

Results. An adaptive user interface model for intelligent systems has been developed, which includes a knowledge base model of the user interface, an agent-oriented model of the user interface, and a library of reusable components that provide integration of the user interface into both individual intelligent systems and intelligent systems groups. A method of transferring user interface components within an intelligent systems group during the operation of an intelligent system has also been developed.

Conclusion. Developed models and tools allow to simplify the reuse of user interface components together with knowledge base and problem-solving components in the design and development of individual intelligent systems, as well as in the design and development of a group of semantically compatible intelligent systems, ensuring automation of integration of user interfaces and their adaptation for each user. The developed set of user interface components has been included in a library of reusable user interface components. Further expansion of the set of components in the library and their integration into a group of semantically compatible intelligent systems is planned.

Objectives. Specifications of models and tools for the development of artificial neural networks (ANNs) and their integration into knowledge bases (KBs) of intelligent systems are being developed. The relevance is determined by the necessity of implementing the possibility to solve complex problems by intelligent systems, which algorithms and methods of solving are not available in the knowledge base of the intelligent system.

Methods. Four levels of integration of artificial neural networks into knowledge bases are formulated and analyzed. During the analysis the requirements and specifications for required models and tools for the development and integration are formulated. Specified at each level the models and tools include the models and tools of previous level. The application of the tools is considered by the example of solving the problem of classifying the knowledge base entities using a graph neural network.

Results. The specifications of the ANN representation model in the knowledge base, the agent-based model for the development and interpretation of the ANN, which ensures the integration of the ANN into knowledge bases at all selected levels, as well as the method for classifying knowledge base entities using a graph neural network, have been developed.

Conclusion. The developed models and tools allow integrating any trained ANNs into the knowledge base of the intelligent system and using them to solve complex problems within the framework of OSTIS technology. It also becomes possible to design and train ANNs both on the basis of external data and on the basis of fragments of the knowledge base. Automation of ANNs development process in the knowledge base becomes available.

Objectives. The problem of IT diagnostics of signs of Parkinson's disease is solved by analyzing changes in the voice and slowing down the movement of patients. The urgency of the task is associated with the need for early diagnosis of the disease. A method of complex recognition of Parkinson's disease using machine learning is proposed, based on markers of voice analysis and changes in the patient's movements on known data sets.

Methods. The time-frequency function (the wavelet function) and the Meyer kepstral coefficient function, the KNN algorithm (k-Nearest Neighbors, KNN) and the algorithm of a two-layer neural network are used for training and testing on publicly available datasets on speech changes and motion retardation in Parkinson's disease. A Bayesian optimizer is also used to improve the hyperparameters of the KNN algorithm.

Results. The KNN algorithm was used for speech recognition of patients, the test accuracy of 94.7% was achieved in the diagnosis of Parkinson's disease by voice change. The Bayesian neural network algorithm was applied to recognize the slowing down of the patients' movements, it gave a test accuracy of 96.2% for the diagnosis of Parkinson's disease.

Conclusion. The obtained results of recognition of signs of Parkinson's disease are close to the world level. On the same set of data on speech changes of patients, one of the best indicators of foreign studies is 95.8%. On the same set of data on motion deceleration, one of the best indicators of foreign researchers is 98.8%. The proposed author's technique is intended for use in the subsystem of IT diagnostics of neurological diseases of a Smart city.