Vol 25, No 9 (2019)

The expansion of modular representations to a set of rational fractions with a fixed denominator is analyzed. Methods have been developed for performing basic computer operations for such representations in the bit grid of a computational reconfigurable system. Given in pseudo-language and justified algorithms for their implementation, obtained estimates of their temporal and tabular complexities. The expediency of using a quadratic modular system to reduce the complexity of individual iterative procedures that underlie operations on the introduced modular representations and which have quadratic complexity has been proved. The representations introduced are intended for use in modular reconfigurable computing SIMD systems.

The article is devoted to the study of the G/G/1 type HE₂/H₂/1 queuing system with a second-order hypererlangian input distribution and a hyperexponential service time law with the aim of obtaining a solution for the average waiting time in queue in the case of stationary mode. For this, the classical method of spectral decomposition of the solution of the Lindley integral equation is used. For practical application of the obtained results, the method of moments is used. It turns out that the hyperelangian distribution law HE₂, like the hyperexponential H₂, which is three-parameter, can be determined by both the first two moments and the first three moments. The choice of such probability distribution laws is due to the fact that they are the most common distributions of non-negative continuous random variables, since the coefficient of variation for the HE₂ c_τ ≥ 1 / $\sqrt{2}$ distribution covers a wider range than the hyperexponential distribution for which c_τ ≥ 1. Determination of the principal characteristic of QS G/G/1 — of the average waiting time in queue an important task due to the fact that for such a QS there is no solution in the general case. The method of spectral decomposition of the solution of the Lindley integral equation for the QS HE₂/H₂/1 allows one to obtain a solution in closed form.

In this paper we focus on the problem of user preferences prediction using the gallery of his mobile device. We consider such categories of interests as interior items, food, transport and sport equipment. The novel two-phased method has been proposed. At the first stage, the facial regions are detected on all photos and videos, and the feature vectors are extracted using deep convolutional neural networks. These feature vectors are grouped using known agglomerative clustering techniques. Finally, we select public photos and videos which do not contain faces from the large clusters. At the second stage, these public images are processed on the remote server using high precision Faster R-CNN object detectors. Objects from other images (personal images) are detected on mobile device in offline mode using SSDLite and MobileNet. In the experimental study several neural network-based detectors have been trained using the united training sample from MS Coco, ImageNet and Open Images datasets. Their comparative analysis demonstrated that the Faster R-CNN-based models are characterized with 30 % higher recall when compared to the SSDLite detectors. However, the latter models process each image 3—9-times faster. Finally, we presented the experimental results of facial clustering with GFW (Grouping Faces in the Wild) dataset using either existing feature descriptors (VGGFace, VGGFace2) or the preliminarily trained MobileNet. The latter model with average link hierarchical clustering achieved the highest B-cubed F-measure.

The article describes the formalization of the threshold method of modular separation of secret using minimally redundant coding. In particular, the conditions were determined to ensure the possibility of using minimum redundant modular arithmetic (MRMA) as the computer-arithmetic basis for the use in threshold cryptographic schemes of secret separation. In comparison with no redundant analogues, the MRMA allows to minimize the complexity of the operation of restoration of the secret-original on the basis of the share secrets belonging to groups of subscribers.

The conditions for the correctness of the threshold principle in the framework of modular coding are also obtained, and on this basis a method has been developed for neutralizing critical situations with ensuring an adequate level of cryptographic resistance.

For project-oriented IT companies the functional areas of process automation and software classes for such automation in these areas are defined in the paper. The overview of most popular software for project and project portfolio management, risk management and task tracker systems is presented. The functional capabilities for project management automation are discussed for such software as ERP-systems SAP S/4 HANA, Microsoft Dynamics 365 for Finance and Operations, 1С:ERP + PM Управление проектной организацией 2 solution. The benefits of ERP-systems for IT company corporate project management system design are revealed. The software structure and corresponding data interaction are proposed.