This paper presents a time-saving method for monitoring the ecology of a dispersed territory using the delivery of measurement units provided by unmanned aerial vehicles with measurement sensors according to a reverse priority algorithm. It is achievable because of the decreasing mean waiting...
30 September 2021 in Informatyka, Automatyka, Pomiary w Gospodarce i Ochronie Środowiska
W artykule przedstawiono wyniki badań poruszających problem poprawy wiarygodności oznaczania stężenia zanieczyszczeń w obiektach biologicznych w cieczy metodą wizualizacji wyładowań gazowych. Opracowano ulepszoną metodę analizy obiektów biologicznych w cieczy opartą na wizualizacji wyładowań gazowych (GDV), zaproponowano podejście kryterialne do oceny składu obiektów biologicznych w cieczy z zastosowaniem tej metody, przedstawiono ocenę charakteru obiektów biologicznych w cieczy wykorzystującą intensywność składowych spektralnych jej promieniowania uzyskanego podczas GDV.Opracowano i zbadano model matematyczny zapłonu wyładowania koronowego oraz zależność intensywności widma promieniowania obiektu biologicznego w fazie ciekłej od jego składu chemicznego, zaproponowano funkcję przeliczeniową do oceny koncentracji zanieczyszczeń wraz z parametrami informacyjnymi obrazów GDV. W artykule zostały przedstawione wszystkie wyniki badań eksperymentalnych GDV i składu spektralnego obiektów biologicznych w fazie ciekłej (LPBO).Proponowane podejście pozwala określić zakres stężeń Mg w płynie z ust, przy różnych schorzeniach tarczycy uzyskanych metodą trylonometryczną.Stwierdzono, że stężenie Mg w płynie z ust pacjentów bez chorób tarczycy wynosi 12,73 ± 2,16 mg/l, u pacjentów z czynnikami ryzyka chorób tarczycy stężenie wynosi 14,98 ± 1,92 mg/l, u pacjentów z wolem stężenie wynosi 26,65 ± 3,73 mg/l.Dane te pozwalają na zapewnienie pacjentom lepszej diagnostyki zaburzeń patologicznych w tyreocytach gruczołowych, które opierają się na stężeniu Mg w płynie ustnym.Potwierdza się, że stężenie Mg w płynie ustnym większe niż 15 mg/l może wskazywać na obecność patologii trylonometrycznej, w tym ogniska w tarczycy.
In this article, we discuss an extensive class of channel codes called turbo codes. These error correction methods will achieve very good results in terms of error rates, which may be close to the bandwidth limit of the Shannon channel. The article begins with a brief discussion on the coding of the turbo, and then describes the form of the iterative decoder most often used to decode the turbo codes. This article proposes a new optimal modification of the log-MAP-log decoding algorithm. This method (PL-log-MAP) is based on a partial linear approximation of the correction function in the Jacobean logarithm. Using the proposed approximation, the complex functions of ln (.) I exp (.) In the log-MAP algorithm can be estimated with high accuracy and lower computational complexity. The effectiveness of the proposed approximation is tested and demonstrated by applying it to the digital communication system for transmission images in MatLab. It seems that the performance of the PL-log-MAP algorithm has the closest efficiency to the original log-MAP solution.
This paper discusses the promising modern turbo-codes in the concept of 3 rd generation communication systems. Presents the extended mathematical principles of the algebra logarithm of the likelihood function for N statistically independent information bits with the transition to trigonometric and hyperbolic functions by using Gudermannian expression and decomposition in B. Taylor power series for finding the result with the defined accuracy.
27 December 2024 in Natural & Mathematical Sciences in Medicine and Medical Education
An analysis of existing cardiovascular system models with regulatory loops used for research purposes and disease diagnosis in medicine has been conducted. A dynamic model of the cardiovascular system that accounts for regulatory processes is proposed. The model incorporates one of the key aspects of blood circulation regulation—neural regulation—achieved through the sympathetic and parasympathetic nervous systems. The sympathetic nervous system stimulates the heart to increase the rate and strength of contractions and causes the constriction of peripheral vessels, thereby raising blood pressure. Conversely, the parasympathetic nervous system decreases the heart rate and promotes vessel dilation, lowering blood pressure. These mechanisms interact closely to maintain stable blood circulation in response to the organism's changing physiological needs. Humoral regulation includes the action of various hormones and bioactive substances circulating in the blood that affect the cardiovascular system, which is also considered in the model. The dynamic component of the regulatory system at the intermediate level (humoral system) in the model simulates the formation of mediators, such as adrenaline and noradrenaline, which travel through the blood vessels. The effects of hormones from the intermediate level on the cardiovascular system are sensed by receptors, such as baroreceptors and chemoreceptors, located in the aorta and the pulmonary circulation, among other places. To enhance the model's functionality, the intermediate regulatory level of the humoral system incorporates a Kolmogorov-Arnold Networks (KANs) system. The KAN network is trained on a knowledge base derived from dozens of acute critical cardiovascular situations. This model can subsequently be used for computer-based prediction and diagnosis of patient diseases and for training medical students.
5 October 2026 in Інфокомунікаційні та комп’ютерні технології
The article investigates the problem of network traffic anomaly detection in microservices architectures, which are characterized by high dynamism, numerous short-lived connections, and complex inter-service communication patterns. It is shown that traditional anomaly detection approaches based on signature analysis and static rules are insufficiently effective in such environments due to the high variability of traffic behavior.This study aims to evaluate the effectiveness of machine learning models in detecting network traffic anomalies, taking into account the specific features of microservice systems. Within the scope of the research, the problem is formulated as a binary classification task in which each network flow is assigned to either a normal or an anomalous class. The CIC-IDS2017 dataset, which contains over 2 million network traffic records across various attack types, is used for experimental evaluation.Feature selection is performed based on the characteristics of the microservices architecture, including traffic intensity, session duration, statistical properties of inter-packet intervals, and indicators of traffic flow variability. Three machine learning models are implemented and evaluated: Isolation Forest, One-Class SVM, and an autoencoder. Model performance is assessed using Precision, Recall, and F1-score metrics. The experimental results demonstrate that all models achieve satisfactory accuracy; however, the autoencoder performs best, achieving the highest values across the main evaluation metrics. This indicates the effectiveness of deep learning methods in detecting complex nonlinear patterns in network traffic.The obtained results confirm the feasibility of applying machine learning models to improve the efficiency of security monitoring systems in distributed information environments. The use of intelligent analysis methods enables the timely detection of anomalous deviations in network traffic that may indicate cyber threats or system malfunctions