Abstract: This paper examines the traffic rules education and examination system in Hungary, by using the Quality Function Deployment (QFD) method, as a new approach towards this complex topic. The education and examination of traffic rules are necessary for the stakeholders, but they have slightly different goals and objectives. This system has two separate stakeholders, the citizens and the authority, with their own set of goals, objectives, desires and ideas, about this system. The QFD reveals the connections between these layers. The paper analyses statistical data regarding road safety and presents the QFD model of both stakeholders and their inter-connections. The results of this work can be used to redesign education and examination methods, during the application of digitalized egovernment solutions and as a general approach to match individual and public interests.

Keywords: Quality Function Deployment; Traffic rules; Traffic safety; E-government; E-learning

Abstract: Diabetic retinopathy (DR), is currently one of the major causes of preventable blindness, worldwide. With an early diagnosis and proper treatment of this eye disease, we can prevent the spread of diabetic retinopathy. In this paper, we propose a new alternative of local binary convolutional neural network (LBCNN) deterministic filter generation which can approximate the performance of the standard convolutional neural network (CNN) with less learnable parameters and also with less memory use, which can be helpful in systems with low-memory or low computational capacity, like smart-phones. We compare our scheme with standard CNN and LBCNN that uses stochastic filter generation strategy on retinal fundus image datasets in case of binary classification into healthy and damaged classes. These experiments are also evaluated according to the standard criteria used in medical applications, such as, overall accuracy, specificity, sensitivity and predictive values. On the small dataset (Aptos), one of our proposed LBCNN architectures outperformed all of the other deep learning models examined.

Keywords: CAD (Computer-aided diagnostics); Binary classification; Memory reduction; Learnable parameters

Abstract: In this paper, a novel framework is introduced by combining compressive sensing(CS) theory, digital curvelet transform, and Principal Component Analysis to improve the performance of face recognition method. CS is a highly attractive approach in the field of signal processing, which provides an efficient way of data sampling at a lower rate than the Nyquist sampling rate. CS offers numerous advantages, like less memory storage, less power consumption and higher data transmission rate etc. Here, CS is used on the face images, which offers reduction in storage space and computational time. The use of curvelet transform provides dual benefits: (i) sparse representation (ii) improvement on detailed content. To extract the feature vector, the Principal Component Analysis is then applied. The Performance of the proposed face recognition method is computed by applying cross-validation technique, compressive sensing based classifier, neural network, Naive Bayes and Support Vector Machine classifier. The proposed technique can efficiently perform the face recognition, at a low computational cost. Extensive experiments, on ORL and AR face databases, are conducted to validate our claim. The proposed technique also recognizes face images more efficiently than the traditional PCA, with a 1.5% higher recognition rate, if a person wears a face mask, as protection from COVID-19

Keywords: Face Recognition; Compressive Sensing; COVID-19; Curvelet Transform

Abstract: The main goal of our research, and thus, of our present study, was to explore some problems and issues of business behavior and etiquette in Slovakia and Hungary. The international comparative research program launched by Fam and Richards was our starting point, in which we examined these two countries. We found that due to the cultural differences in the dimensions of the Hofstede model, differences can be detected in business ethics and etiquette in the business life of Hungary and Slovakia, which can be supported by statistical methods. At the same time, our results also showed that almost a half-century since Hofstede research has not passed without a trace in the Central European Region. The transition from socialism to a market economy involved border openings. At the same time, it facilitated the convergence of the business culture of Slovakia and Hungary, changing the relative position of these two countries on the Hofstede scale. We drew attention to the fact that it would be worth repeating Hofstede's research to record socioeconomic changes, in the case of intensely transforming societies and countries.

Keywords: business behavior; business etiquette; business ethics; cultural differences

Abstract: This paper presents an efficient approach, that is centered on a chaotic symbiotic organism search (CSOS) algorithm, for solving the energy management optimization (EMO) problem in Micro-grids (MG) containing diverse distributed generation resources (DGR) besides energy storage systems. The proposed approach is equipped with a chaotic map to guarantee a wider coverage of the search space and rapid time for convergence when searching solutions for the EMO problem under the various exploiting constraints. The CSOS approach is examined on a practical microgrid linked to public services. The effectiveness of CSOS is proven through a comparison of the obtained solutions, in terms of operating costs, with those of other scalable algorithms, such as, GA and PSO.

Keywords: Micro Grids; Energy Management Optimization; Distributed Generation Resources, Chaotic Symbiotic Organism search algorithm

Abstract: The efficiency of organizational processes largely depends on the quality of Knowledge Management. In the crisis situation, caused by the Coronavirus, its significance is especially apparent. Our research sought to reveal what knowledge management problems have emerged, due to the spread of the Coronavirus and what difficulties need to be coped with, in organizations. During the research, in three small groups within an online workshop, a group of experts worked with the “Be-novative” program, which uses design thinking, to connect the innovation process with the power of gamification and crowdsourcing. Using the program, ideas were formulated through joined-up thinking, evaluated online and further developed. Out of the 141 problems/ideas raised, based on the community’s evaluation, three complex solution possibilities were developed, which combine several ideas under comprehensive titles. The developed proposals were published on the website of the professional organization, thus, supporting the successful functioning of knowledge management programs.

Keywords: Be-novative; Covid-19; knowledge management; online

Abstract: In this article, we review recently proposed, advanced methods, for the control of blood glucose levels, in patients with type 1 diabetes. The proposed methods are based on various techniques, such as predictive control, filters, and machine learning. Results have shown that the artificial pancreas may control blood glucose levels better than conservative insulin administration, while avoiding the risk of hypoglycemia or hyperglycemia. The most commonly used methods for controlling blood glucose levels are giving good results, while methods based on machine learning algorithms also offer promising performance. Nevertheless, there are numerous challenges in designing algorithms for the artificial pancreas, which need to be considered. The aim of this research is to provide an overview of the latest achievements in this research field, find the best solutions and, ultimately, improve them in the future.

Keywords: Artificial pancreas; continuous glucose monitoring; model predictive control; sliding mode control; Kalman filters; machine learning; neural networks; type 1 diabetes

Abstract: This paper deals with the application of artificial immune networks in continuous function optimizations. The performance of the immunological algorithms is analyzed using the Optimization Algorithm Toolkit. It is shown that the CLIGA algorithm has, by far, the fastest convergence and the best score - in terms of the number of required iterations, for the analyzed continuous function. Also, based on the test results, it was concluded, that the lowest total number of iterations for the defined run time was achieved with the opt-IA algorithm, followed by the CLONALG and CLIGA algorithms.

Keywords: artificial immune networks; Optimization Algorithm Toolkit; continuous function optimization; performance

Abstract: Selective Waste Collection is an essential part of recycling raw materials, in order to protect our environment. The waste collection is carried out by using a seriously polluting vehicle, due to the fact that most of the gathering vehicles are using fossil energy sources, like gasoline. High volume of carcinogenic elements are contained in the emitted exhaust gases. The current waste collection methods are just focusing on the load of the selective waste collective vehicle during a collection route. The goal of this research is to find the best solution to use the full storage capacity of the selective waste collecting vehicle with the lowest volume of residual air due to the effectively compressed waste. The closed and uncompressed PET bottles require the largest volume in the waste collecting vehicle. It is essential to minimize the air in the PET bottles to decrease the volume. Different methods have been examined to increase the density of the selectively collected waste. Statistical data have been used to determine the collecting parameters - nowadays, a 15 t waste collecting vehicle, with a 20 m3 load capacity is only gathering 1-1.5 t PET due to the ineffective use of its load compartment. The application of the method herein, enhances the efficiency of the waste collecting vehicle by gathering 4-10 times more waste than currently used methods.

Keywords: selective waste; waste gathering; capacity utilization

Abstract: This paper focuses on malware analysis and detection using machine learning methods. The aim of the authors was to perform static and dynamic analysis of programs designed for Windows and then to present the results of the analysis as a dataset. We analysed and implemented different classification methods, such as decision trees, random forests, support vectors and naive Bayes methods. We verified their ability to distinguish malicious and harmless samples and evaluated their success rate using classification accuracy metrics. Then, we compared the results obtained by prediction over the dataset generated by static and dynamic analysis. Classification was more successful on the data gained using the dynamic analysis method. The best malware detection algorithms have been found to be decision tree-based algorithms, in particular the random forest algorithm, which achieves excellent malware detection accuracy of up to 95.95% with a standard deviation of only 0.58%.

Keywords: malware; static analysis; dynamic analysis; dataset; classification

Abstract: In this paper a new type of passive mechanism for vibration suppression is introduced. The mechanism is based on the system of cantilever - mass units (dynamic absorbers) connected to the basic structure. The support of cantilevers is rigid or even elastic. The vibration of the system is caused by external excitation force which acts on the basic structure. The aim of the paper is to determine the parameters of the system for which the frequency gap and the vibration suppression occur. The used mathematical model is a system of coupled equations where the measured parameters are introduced by the application of a newly developed, so-called, ‘elastic support method’. Solving the mathematical model, the amplitude-frequency vibration property of the system is obtained. The computed solution is compared with that the previously published result of the ‘wallpaper’ type metastructure for vibration suppression, which is modeled as a system of translation moving system of mass-in-mass units. It is concluded that the effect of the suggested mechanism is in good agreement with that of the metastructure for vibration suppression. The resonances of the two models are matched with the results of Inventor Finite Element Analysis, too. Difference in results is negligible.

Keywords: "wallpaper"-like metamaterial; cantilever-mass mechanism; vibration suppression; 5-DoF system; natural-frequency

Abstract: In computer networks, intrusion detection systems are used to detect cyber-attacks and anomalies. Feature selection is important for intrusion detection systems to scan the network quickly and accurately. On the other hand, analyzes performed using data with many attributes cause significant resource and time loss. In this study, unlike the literature studies, the frequency effects of the features in the data set are analyzed in detecting cyber-attacks on computer networks. Firstly, the frequencies of the features in the NSL-KDD data set were determined. Then, the effect of high-frequency features in detecting cyber-attacks has been examined with the widely used machine learning algorithms of Random Forest, J48, Naive Bayes, and Multi-Layer Perceptron. The performance of each algorithm is evaluated by considering Precision, False Positive Rate, Accuracy, and True Positive Rate statistics. Detection performances of different types of cyberattacks in the NSL-KDD dataset were analyzed with machine learning algorithms. Precision, Receiver Operator Characteristic, F1 score, recall, and accuracy statistics were chosen as success criteria of machine learning algorithms in attack detection. The results showed that features with high frequency are effective in detecting attacks.

Keywords: Attribute selection;Cyberattacks; Machine Learning; IDS;NSL-KDD; Anomaly detection

Abstract: Solder joint reliability is critical in the design of advanced microelectronic packaging. Predictions of reliability by thermo-mechanical simulations can accelerate the evaluations of advanced packaging and the introduction of novel solder materials. In this work, a finite element model of a thermally loaded Fan-Out Wafer Level Package (FO-WLP) was built and analyzed focusing on the creep behavior of the solder balls and the consequent effect on the reliability of the package. The lead-free soldering materials in the analyses were either of a widely used SAC305, or novel doped SAC solders as SAC-R, SAC-Q and InnoLot. Visco-plastic (Anand creep) properties for the solders were defined as study parameters, where 6 variations were used for the described SAC305, and further 3 sets for the doped SAC solders, respectively. Identifying a stress concentration at the sharp bond pad edges by modeling ideal geometries, a refined geometry was introduced and evaluated. Simulations for a 3-cycle thermal load were conducted, and results were collected and analyzed for Creep Strain and Strain Energies in critical positions in the solder, and reliability prediction was performed based on Morrow’s model. Results show the benefit of the refined compositions of Doped SAC solders on the mechanical behavior and improved reliability.

Keywords: Reliability; Creep behavior; lead-free Solder