In this paper, we propose Federated Deep Learning (FDL) for intrusion detection in heterogeneous networks. Local Deep Neural Network (DNN) models are used to learn the hierarchical representations of the private network traffic data in multiple edge nodes. A dedicated central server receives the parameters of the local DNN models from the edge nodes, and it aggregates them to produce an FDL model using the Fed+ fusion algorithm. Simulation results show that the FDL model achieved an accuracy of 99.27 ± 0.79%, a precision of 97.03 ± 4.22%, a recall of 98.06 ± 1.72%, an F1 score of 97.50 ± 2.55%, and a False Positive Rate (FPR) of 2.40 ± 2.47%. The classification performance and the generalisation ability of the FDL model are better than those of the local DNN models. The Fed+ algorithm outperformed two state-of-the-art fusion algorithms, namely federated averaging (FedAvg) and Coordinate Median (CM). Therefore, the DNN-Fed+ model is preferable for intrusion detection in heterogeneous wireless networks.

Accurate downlink channel state information (CSI) is one of the essential requirements for harnessing the potential advantages of frequency-division duplexing (FDD) massive multi-input multi-output (MIMO) systems. The current state-of-art in this vibrant research area include the use of deep learning to compress and feedback downlink CSI at the user equipments (UEs). These approaches focus mainly on achieving CSI feedback with high reconstruction performance and low complexity, but at the expense of inflexible compression rate (CR). High training overheads and limited storage capacity requirements are some of the challenges associated with the design of dynamic CR, which instantaneously adapt to propagation environment. This paper applies transfer learning (TL) to develop a multi-rate CSI compression and recovery neural network (TL-MRNet) with reduced training overheads. Simulation results are presented to validate the superiority of the proposed TL-MRNet over traditional methods in terms of normalized mean square error and cosine similarity.

Intelligent reflecting surface (IRS)-aided millimeter-wave (mmWave) multiple-input single-output (MISO) is considered one of the promising techniques in next-generation wireless communication. However, existing beamforming methods for IRS-aided mm Wave MISO systems require high computational power, so it cannot be widely used. In this paper, we combine an unsupervised learning-based fast beamforming method with IRS-aided MISO systems, to significantly reduce the computational complexity of this system. Specifically, a new beamforming design method is proposed by adopting the feature fusion means in unsupervised learning. By designing a specific loss function, the beamforming can be obtained to make the spectrum more efficient, and the complexity is lower than that of the existing algorithms. Simulation results show that the proposed beamforming method can effectively reduce the computational complexity while obtaining relatively good performance results.

Malware traffic classification (MTC) is a key technology for solving anomaly detection and intrusion detection problems. And hence it plays an important role in the field of network security. Traditional MTC methods based on port, payload and statistic depend on the manual-designed features, which have low accuracy. Recently, deep learning methods have attracted significant attention due to their high accuracy in terms of classification. However, in practical application scenarios, deep learning methods require a large amount of labeled samples for training, while the available labeled samples for training are very rare. Furthermore, the preparation of a large amount of labeled samples requires a lot of labor costs. To solve these problems, this paper proposes two methods based on semi-supervised learning (SSL) and transfer learning (TL), respectively. Our proposed methods use a large amount of unlabeled data collected in the Internet traffic, which can greatly improve the accuracy classification with few labeled samples. Through experiments, we obtained the best method to improve the accuracy of few labeled samples in different situations. Experiment results show that our proposed methods can satisfy the requirement of MTC in the case of few labeled samples.

: In this paper, the influence of MgO and CaO content on the quality of technical ceramics (which originally represents the Al 2 O 3 - SiO 2 - CaO - MgO system) has been investigated. Therefore, quality tests were performed on samples where the contents of CaO and MgO were taken as variable values. Based on the obtained test results and their analysis, certain characteristics of ceramics are defined with the different percentages of individual oxides content in 98.2 - 99% Al 2 O 3 .

This research paper deals with the problem of Metal-Oxide Surge Arrester (MOSA) condition monitoring and a new methodology in surge arrester monitoring and diagnostics is presented. A machine learning algorithm (back propagation regression) is used to estimate the non-linearity coefficient of the surge arrester, based on operating voltage and leakage current of the arrester. Using a simulated system, this research investigates the possibility of application and efficiency of machine learning. It is shown that the applied learning algorithm results are competitive with the model results parameters calculated as R2 = 0.999 and mean absolute real error computed as 0.005 which has shown that the proposed model can be used for MOSA monitoring and diagnostic purposes.

Glioblastoma multiforme (GBM) is the most frequent type of primary astrocytomas. We examined the association between single nucleotide polymorphisms (SNPs) in Aurora kinase A (AURKA), Aurora kinase B (AURKB), Aurora kinase C (AURKC) and Polo-like kinase 1 (PLK1) mitotic checkpoint genes and GBM risk by qPCR genotyping. In silico analysis was performed to evaluate effects of polymorphic biological sequences on protein binding motifs. Chi-square and Fisher statistics revealed a significant difference in genotypes frequencies between GBM patients and controls for AURKB rs2289590 variant (p = 0.038). Association with decreased GBM risk was demonstrated for AURKB rs2289590 AC genotype (OR = 0.54; 95% CI = 0.33–0.88; p = 0.015). Furthermore, AURKC rs11084490 CG genotype was associated with lower GBM risk (OR = 0.57; 95% CI = 0.34–0.95; p = 0.031). Bioinformatic analysis of rs2289590 polymorphic region identified additional binding site for the Yin-Yang 1 (YY1) transcription factor in the presence of C allele. Our results indicated that rs2289590 in AURKB and rs11084490 in AURKC were associated with a reduced GBM risk. The present study was performed on a less numerous but ethnically homogeneous population. Hence, future investigations in larger and multiethnic groups are needed to strengthen these results.

Prevention of cardiovascular events and regression of atherosclerotic changes are the primary aims of preventive cardiovascular medicine. Arterial thrombosis is caused by endothelial dysfunction, which disrupts vascular haemostasis. Glucagon-like peptide 1 (GLP-1) receptor agonists have been initially used as glucose lowering agents, but over time have been used for other indications due to their cardiorenal benefit, as well as their benefit in the regression of atherosclerosis process. The aim of this paper is to present the benefits of GLP-1 receptor agonists in the prevention of atherosclerotic changes, in the preservation of brain vascular function, and to show the possible role in the treatment of neurodegenerative diseases.

Nowadays there are ham and spam messages that are sent to the users via SMS. The aim of this article is to show how machine learning and text processing technologies can be used in order to predict the trustworthiness of SMS messages. The data we are going to use is collected from Kaggle. This study is very important because it helps us to understand how machine learning and text processing can be used in order to predict message trustworthiness. At the time of writing this article, there was not an article explaining how this can be done using the Multinomial Naive Bayes algorithm. The methodology we used in this article consists of dataset collection, data cleaning, data analysis, text preparation, and training model. This will be seen in the methodology section in great detail. At the end of this article, we will show to u the accuracy that we have got when implementing a Multinomial Naive Bayes algorithm for the classification of SMS messages. This study was quite beneficial because anyone can see how Multinomial Naive Bayes algorithm usage can be beneficial in order to predict the trustworthiness of SMS messages.

Water quality is deteriorating over the years, and the main source of water pollution is industrial, agricultural and municipal wastewater. Heavy metals, organic compounds and microorganisms, present even in traces, can be very dangerous to human health, aquatic organisms and the environment. Therefore, in this study was investigate the possibility of modified and unmodified plum pits as biosorbents for Pb (II) ions removal from aqueous solution. Experimental data have shown that these bisorbents show a certain potential for application in the metal removal process. The feasibility was tested for an unmodified and modified biosorbent based on plum pits in the range of concentrations 150-200 mg/l (unmodified sample) and 100-200 mg/l (modified sample) at a contact time of 30 and 60 minutes . Adsorption parameters were determined using the Freundlich isotherm. The results showed that unmodified biosorbent based on plum pits with increasing concentration from 150 mg/L to 200 mg/L leads to a large increase in the percentage of removal of Pb (II) ions, with no significant effect on contact time. In contrast to the unmodified sample, the modified biosorbent based on plum pits % of removed Pb (II) ions significantly increases the contact time at the initial Pb (II) concentration of 100 mg/L, while at the initial concentration of 150 mg/L and longer mixing, the removal efficiency increases and amounts to 86.032 %. The calculated values of the parameters used in the Freundlich isotherm indicated the existence of high-energy sorption centers in the unmodified bisorbent based on plum pits, while the calculated values of the parameters used in the Freundlich isotherm for the modified biosorbent based on plum pits showed moderate mode adsorption.

: Construction materials in the form of any products are subject to unintentional or harmful changes, occurrences and processes that reduce their usability. The destruction of construction materials is aimed to be slowed down or prevented by measures and procedures of a special technological discipline - material protection, which is usually called surface protection, since harmful occurrences and processes mostly begin on the surface of the product. In addition to the many protective methods that are used, corrosion inhibitors have a special place due to their specificity and widespread use. Based on the performed tests and their analysis, it was determined that the inhibition efficiency obtained by electrochemical measurements is in good correlation with the results obtained by the FTIR method. Impedance measurements of steel St 37-4 Pectin C in the tested media show corrosion resistance. Pectin C in 3.5% HCl at a concentration of 2.0 g / l increases the value of the charge transfer resistance and the increases of the size of the absolute impedance in the Bode diagram, which further confirms the improved resistance to corrosion of steel.

In this study, metal complex of Copper(II) with a Schiff base derived from 2,2-dihydroxyindane-1,3-dione and 2-aminoethanoic acid were synthesized. The product are characterized by spectral methods. The antimicrobial activity was tested on reference bacterial strains and the antioxidant capacity was analyzed by using the DPPH and FRAP methods. The spectral data indicates that the Schiff base coordinates the Copper(II) as a tridentate ONO donor ligand. The compounds showed weaker antimicrobial activity on certain tested microorganisms. In vitro testing of antioxidant activity showed a significant reducing ability of the complex, as well as inhibitory activity against DPPH radicals.

In this study, the chemical profiles, antioxidant and antibacterial activity of Helivhrysum italicum essential oils from three plantation fields in Herzegovina were analysed. GC/MS analysis showed that all samples were rich in sesquiterpenes (45.19%-50.07%) and monoterpenes (21.15%-23.21%), followed by oxygenated monoterpenes (9.92%-14.03%). Diketones in the essential oil were detected in quantities ranging 5.72% to 6.67%. The main components in essential oils were γ-curcumene, α-pinene, β-selinene and neril-acetate. All tested essential oils exhibited relatively weak DPPH-scavenging capacity. The antimicrobial activity of the essential oil was assayed by using the disk diffusion method. E. coli was most resistant against all three tested H. italicum essential oils, while moderate inhibitory activity against S. aureus and C. albicans was detected. The L. monocytogenes was the most sensitive where all three tested samples showed inhibitory activity.