Search

Real-world multi-objective optimization problems encounter different types of uncertainty that may affect the quality of solutions. One common type is the stochastic noise that contaminates the objective functions. Another type of uncertainty is the different forms of dynamism including changes in the objective functions. Although related work in the literature targets only a single type, in this paper, we study Dynamic Multi-objective Optimization problems (DMOPs) contaminated with stochastic noises by dealing with the two types of uncertainty simultaneously. In such problems, handling uncertainty becomes a critical issue since the evolutionary process should be able to distinguish between changes that come from noise and real environmental changes that resulted from different forms of dynamism. To study both noisy and dynamic environments, we propose a flexible mechanism to incorporate noise into the DMOPs. Two novel techniques called Multi-Sensor Detection Mechanism (MSD) and Welford-Based Detection Mechanism (WBD) are proposed to differentiate between real change points and noise points. The proposed techniques are incorporated into a set of Dynamic Multi-objective Evolutionary Algorithms (DMOEAs) to analyze their impact. Our empirical study reveals the effectiveness of the proposed techniques for isolating noise from real dynamic changes and diminishing the noise effect on performance.

IEEEWhile the zeitgeist of today prescribes localization as a key to eliminate the excessive transportation for manufacturing firms, firm-level practical and workable guidelines that go beyond maxims are scant. Laden with analytical methods, this paper offers a phronetic, customizable framework based on Lean Six Sigma (LSS). The framework mimics Define-Measure-Analyze-Improve-Control (DMAIC) process. The early phases of the framework, ex-ante, prescribe the use of a variety tools and techniques from Lean Six sigma (LSS) and employ fuzzy analytic hierarchy process (FAHP) to extract the vital few root-causes that may impede localization. The latter phases suggest the use of design of experiments (DOE) to evaluate the root-causes, ex-post. The paper outlines a clear and established procedure and demonstrates each analytical phase using evidence from a two-year real-world application in steel industry.

© 2023, The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature.Many prediction methods proposed in the literature can be concerned under two main headings: probabilistic and non-probabilistic methods. In particular, as a kind of non-probabilistic model, artificial neural networks (ANNs), having different properties, have been commonly and effectively used in the literature. Some ANNs operate the additive aggregation function in the structure of their neuron models, while others employ the multiplicative aggregation function. Recently proposed dendritic neural networks also have both additional and multiplicative neuron models. The prediction performance of such an artificial neural network will inevitably be negatively affected by the outliers that the time series of interest may contain due to the neuron model in its structure. This study, for the training of a dendritic neural network, presents a robust learning algorithm. The presented robust algorithm is the first for the training of DNM in the literature as far as is known and uses Huber\"s loss function as the fitness function. The iterative process of the robust learning algorithm is carried out by particle swarm optimization. The productivity and efficiency of the suggested learning algorithm were evaluated by analysing different real-life time series. All analyses were performed with original and contaminated data sets under different scenarios. The R-DNM has the best performance for the original data sets with a value of 2.95% in the ABC time series, while the FTSE showed the best performance in approximately 27% and the second best in 33% of all analyses. The proposed R-DNM has been the least affected by outliers in almost all scenarios for contaminated ABC data sets. Moreover, it has been the least affected model by outliers in approximately 71% of the 90 analyses performed for the contaminated FTSE time series. The obtained results show that the dendritic artificial neural network trained by the proposed robust learning algorithm produces the satisfactory predictive results in the analysis of time series with and without outliers.

Two wheeled Self balancing robots have become more popular in recent times and used in many applications, commonly used in transportation as these balancing robots are faster and apparently stable. Balancing is achieved by using controllers tuned for the system. Inverted pendulum principle is used in these robots. This paper proposes the use of two control methods, namely Proportional-Integral-Derivative (PID) and Extended PID (EPID), for maintaining the balance of two-wheeled and self-balancing robots, even when slightly tilted, thus preventing them from falling. PID control theory generally depends up on transfer functions, this paper extends the theory using EPID. EPID provides a new perspective on the working principle of PID control and its relationship with other control theories, such as tracking control and disturbance observer. This approach is illustrated using the fundamental principle of inverted pendulum technology. The key advantage of these robots over other mobile robots is their increased stability due to zero turning radius, which allows for smoother rotations during changes in direction.

© 2023 Elsevier B.V.This paper studies the cluster consensus problem for networks with antagonistic interactions modelled by adjacency matrices with negative weights. By introducing an extended digraph representation consisting of purely cooperative interactions with lifting approach, we relate the trajectories of the system to those of an extended system with a positive digraph. The behaviours of agents in a signed network are extracted from its extended digraph. Consequently, the number of clusters and the cluster members are explicitly determined for any signed digraph by using primary and secondary layer subgraph concepts. The relation between the dynamics of a signed system and its extended representation is investigated for first and higher-order systems. The conditions that make both systems stable are stated. Additionally, the control parameters to achieve cluster consensus are derived explicitly for first, second and third order systems. The obtained results for continuous-time networks are subsequently extended to discrete-time networks. Finally, theoretical results are illustrated via numerical examples.

In this study, forecasting the number of immigrants on the Turkey\"s maritime line for use in a national security project carried out by Turkish Government within the scope of fight against uncontrolled immigration is discussed for the first time. Handling with the immigration problem is one of the biggest concerns of Turkey as unsupervised immigration can adversely affect the demographic and economic structure of the country. Precautions are needed as the short-, medium- and long-term impacts of undetected immigrants on the country\"s ecosystem are unpredictable, but due to the uncertainties inherent in immigration, the cost of using government resources such as patrol vehicles to capture undocumented immigrants can be extremely high. In order to both minimize the expenditure problem and keep immigration under control by providing a proper scan, forecasting the number of immigrants on the maritime line route is seen as an important problem and studied by probabilistic and non-probabilistic models. Since the data for 2020 and 2021 could not be attained yet due to COVID-19, in order to obtain forecasts and compare actual observations for 2019, which is the primarily focus of the research in this study, the dataset of interest on the number of daily immigrants between years 2016 and 2019 is obtained from Turkish Coast Guard Command within Ministry of Interior of Republic of Turkey. To obtain the most accurate forecasts, seven distinguished forecasting methods, from simple to complex, are implemented. Then, the forecast combination approach with meta-fuzzy functions which combines all methods is proposed. Consequently, the forecasting results are acquired and evaluated by using R. The evaluation of the results is made by using widely considered measurement accuracy metric root mean square error. According to the final assessments, the proposed approach gives more accurate forecasting results for the expected number of immigrants on the Turkey\"s maritime line and these results become an input to the national security project.

Smart retail stores started to take place in our lives. Several computer vision and sensor-based systems are working together to achieve such a complex and automated operation. Besides, the retail sector has several open and challenging problems which can be solved by embedded computer vision systems. One important problem to be tackled is shelf control or stock out detection. Here, shelves in a store should be controlled regularly such that no item is missing in the shelf. In this study, we propose an embedded computer vision system to solve this problem. To do so, we frame the shelf control operation as change detection. Due to the constraints posed by the retail sector, we formed the system by an ultra-low or low power microcontroller with an embedded camera attached to it. We provided all the implementation details of the system both from software and hardware perspectives. We also tested the proposed shelf control system from different perspectives. Hence, the reader can form such a system for retail sector or for a broad class of change detection problems in which stand-alone embedded system usage is mandatory.

Intuitionistic fuzzy time series models consider observations hesitation degree but they use memberships and non-membership values together as inputs in the prediction system. The usage of membership and non-membership values as inputs in separate prediction models and combining the outputs of these separate models will provide a more flexible computational approach. Thus, different effects of membership and non-membership degrees on the predictions can be revealed. In this paper, an intuitionistic fuzzy time series prediction model (IFTS-PM) has been proposed. The proposed IFTS-PM uses a new hybrid sigma-pi neural network (HSP-NN), introduced for the first time in the literature, to determine nonlinear relationships between inputs and outputs. In addition, this newly proposed HSP-NN has the ability to multiply linear functions of inputs by unequal weights and convert them to nonlinear relationships. The structure of the proposed IFTS-PM consists of three parts. Two different HSP-NNs generate predictions by taking into account the different contribution levels of memberships and non-memberships. The last part is the part where these predictions are combined. Modified particle swarm optimization is performed to obtain optimal weights of HSP-NNs as well as the combination weights. And by taking the advantage of intuitionistic fuzzy C-means, fuzzy clusters, membership and non-membership values of observations are obtained. Performance of the proposed model is verified by applying it on 48 time series data sets. With all used indications, it has been clearly observed that proposed model has produced outstanding predictions compared to some other state-of-the-art prediction tools.

Due to the complex mathematical structures of the models in engineering, heuristic methods which do not require derivative are developed. This paper improves recently developed Grey Wolf Optimization Algorithm by extending it with three new features: namely presenting a new formulation for evaluating the positions of search agents, applying mirroring distance to the variables violating the limits, and proposing a dynamic decision approach for each agent either in exploration or exploitation phases. The performance of Advanced Grey Wolf Optimization (AGWO) method is tested using several optimization test functions and compared to several heuristic algorithms. Moreover, a planning problem in smart grids is solved by considering different objective functions using 33 and 141 bus distribution test systems. From the numerical simulation results, we observe that, AGWO is able to find the best results compared to other methods from 10 and 9 out of 13 test functions for 30 and 60 variables, respectively. Similar to this, it finds best function values for 5 out of 10 fixed number of variable test functions. Also, the result of the CEC-C06 2019 benchmark functions shows that AGWO outperforms 8 for optimization problems from 10. In power distribution system planning problem, better objective function values were determined by using AGWO, resulting a better voltage profile, less losses, and less emission costs compared to solutions obtained by Grey Wolf Optimization (GWO) and Particle Swarm Optimization (PSO) algorithms.

In this paper, we propose an improved face recognition framework where the training is started with a small set of human annotated face images and then new images are incorporated into the training set with minimum human annotation effort. In order to minimize the human annotation effort for new images, the proposed framework combines three different strategies, namely self-paced learning (SPL), active learning (AL), and minimum sparse reconstruction (MSR). As in the recently proposed ASPL framework [1], SPL is used for automatic annotation of easy images, for which the classifiers are highly confident and AL is used to request the help of an expert for annotating difficult or low-confidence images. In this work, we propose to use MSR to subsample the low-confidence images based on diversity using minimum sparse reconstruction in order to further reduce the number of images that require human annotation. Thus, the proposed framework provides an improvement over the recently proposed ASPL framework [1] by employing MSR for eliminating “similar” images from the set selected by AL for human annotation. Experimental results on two large-scale datasets, namely CASIA-WebFace-Sub and CACD show that the proposed method called ASPL-MSR can achieve similar face recognition performance by using significantly less expert-annotated data as compared to the state-of-the-art. In particular, ASPLMSR requires manual annotation of only 36.10% and 54.10% of the data in CACD and CASIA-WebFace-Sub datasets, respectively, to achieve the same face recognition performance as the case when the whole training data is used with ground truth labels. The experimental results indicate that the number of manually annotated samples have been reduced by nearly 4% and 2% on the two datasets as compared to ASPL [1].

\"Yapay Zeka uygulamaları işletmelerin dijital dönüşümleri açısından önemli biraraştırma alanı olmakla beraber ilgili alan yazında kendine pek fazla yer bulamamış,ana tartışmaların dışında kalmıştır. Ancak giderek artan çeşitlilikte uygulamalarınişletmeler tarafından kullanılması, uygulamaların giderek daha da yaygınlaşması vegüvenilirlik düzeylerinin yükselmesi ile daha çok ele alınması gerektiği görülmüştür.Bu bağlamda turizm ve seyahat bürolarının pek çok ihtiyacına cevap veren YZ uygulamalarını sınıflandırmak ve sektör bazlı değerlendirmeler yapabilmeyi hedefleyenyazımızda Raisch ve Krakowski’ye (2021)* atıfla iki yönetim uygulaması paradigmasınıele aldık. Bu iki paradigma üç farklı stratejik anlayış doğurmaktadır. Bunlar, sadeceogmantasyon odaklı YZ uygulamalarına yatırım yapmak, sadece otomasyon odaklıYZ uygulamalarına yatırım yapmak ve Her ikisine de göreceli ağırlık verecek bir yatırım stratejisi izlemek olarak sayılabilir.\"*Raisch, S., Krakowski, S. 2021. \"Artificial intelligence and management: The automation–augmentation paradox\". Academy of Management Review, 46(1), 192-210.

© 2022 IEEE.Businesses are concerned about energy losses. Electronic gadgets have become more prevalent as a result of their adoption. The frequency with which home electricity usage data is collected has grown, allowing for sophisticated data analysis that was previously unavailable. Indeed, adopting Smart Grid (SG) networks, which are freshly improved networks of linked devices, may considerably enhance the existing energy infrastructure\"s dependability, economy, and durability. The SG involves sharing a lot of data, including information on specific users\" power use. And using this information, machine learning and deep learning algorithms may be able to detect power theft users reliably. This paper presented a Convolutional Neural Network (CNN)-based model for automated network-based vulnerability scanning that has excellent classification performance in many categories. Testing from iteration two to four iterations, this study examines research to discover the ideal configuration of the sequential model (SM) for categorization. The method is validated using a two-layer design, including an efficiency of 0.92, the whole first layer is composed of 128 nodes while the second level consists of 64 nodes. This allows for the development of a higher-level classifier for electrical signals, which may be employed in a number of applications. CNN was used to create electrical signal detectors, and SM was used to extract data from an electricity usage dataset. Furthermore, the number of features in the data set can be reduced using the Blue Monkey (BM) approach, and these results are then used to develop high-performance models. In this regard, the focus of this study has been on lowering the amount of needed features in the dataset in order to establish a rising classification algorithm for electrical signals. Experiments have applied the proposed systems\" fantastic performance, with just 666 characteristics required to combine the CNN and BM methods. Comparative to 1035 traits when CNN was used alone. This shows that the CNN and BM models are better than the CNN model in terms of lowering sufficient know while maintaining the same reliability.

© 2022 IEEE.Deep transfer learning is one of the common techniques used to classify different types of cancer. The goal of this research is to focus on and adopt a fast, accurate, suitable, and reliable for classification of colorectal cancer. Digital histology images are adjustable to the application of convolutional neural networks (CNNs) for analysis and classification, due to the sheer size of pixel data present in them. Which can provide a lot of information about colorectal cancer. We used ten different types of pr-trained models with two type method of classification techniques namely (normal classification and k-fold crosse validation) to classify the tumor tissue, we used two different kinds of datasets were these datasets consisting of three classes (normal, low tumors, and high tumors). Among all these eight models of deep transfer learning, the highest accuracy achieved was 96.6% with Darknet53 for 5-Fold and for normal classification the highest results obtained was 98.7% for ResNet50. Moreover, we compared our result with many other papers in stat-of-the-art, the results obtained show clearly the proposed method was outperformed the other papers.

Bu çalışmada EMG sinyalleri üzerinde öznitelikler hesaplanmıştır. Bu öznitelikler ile el hareketlerini sınıflandırmak için derin öğrenme algoritmaları kullanılmıştır. Bir zaman serisi olarak toplanan EMG sinyalleri üzerinde zaman alanında hesaplanan öznitelik vektörleri belirli boyutlarda simetrik matrisler olarak kaydedilmiştir. Yeniden oluşturulan ve resim dosyası formatında kaydedilen veri seti ile Evrişimsel Sinir Ağı eğitilmiştir. Bu eğitim sonucunda tüm veriler ile %93, test verileri ile %79 başarı ile hareket sınıflandırması gerçekleştirilmiştir. In this study, the statistical features are calculated with EMG signals. Deep learning algorithms are used to classify hand gestures with these features. the feature vectors calculated in the time domain using the EMG signals recorded as a time series are stored as symmetric matrices with specific sizes. The data set that is rebuilt and saved as image file format is used to train the convolutional neural network. Thereafter the training, motion classification is carried out with a success rate of 93% for all data and 79% for test data

© 2022 IEEE.Prediction of the time that will be spent on IT support tickets is very important for planning and optimization of IT support services that are usually bound with service level agreements. Predicting completion time of a ticket is a difficult problem, which requires substantial experience and technical expertise if done manually by a human. However, it is possible to automate this task using supervised machine learning models given we have a large amount of labeled data. In this study, we employ supervised machine learning algorithms to predict completion time of tickets for IT support. We use a real-world dataset that includes about 17 thousand tickets. We employ data science approaches to preprocess and transform the input and feed to supervised machine learning algorithms for learning models for ticket completion time prediction. More specifically we use Linear Regression, Decision Trees Regression, Random Forest Regression, Support Vector Machines Regression, and Multiple Regression algorithms. For the evaluation of these supervised models, we use several metrics such as MAE, MSE, and MAPE. Our results show varying success levels with different supervised machine learning algorithms for this difficult task. Among the models we train, the Decision Trees and Random Forest Regression show promising results.

Otonom araçların (OA'lar) en temel özelliği otonom olmalarıdır. Ancak aracın dinamik çalışma ortamı nedeniyle kontrol algoritmaları kesin olmayan, yaklaşık ve güvenilmez kararlar verebilir. Bu nedenle, özellikle tutarlı engellerden kaçınma algoritmaları olmak üzere daha sağlam sürüş algoritmalarının oluşturulmasına ihtiyaç vardır. Bazen, engellerden kaçınmak için aracın tamamen durması gerekir. Bu durumda aracın motor freni kontrolü devreye alınabilir. Bu çalışmada, reostatik frenleme olarak bilinen elektrikli fren sistemi ile otonom kara taşıtlarını etkin bir şekilde frenlemek için bulanık bir model önerilmiştir. Reostatik fren sistemi (RFS) kullanıldığı için, tasarlanan bu modelleme için bulanık denetleyicinin giriş değerleri araç hızı ve zemin kayganlığı, çıkış değeri ise reostat direnç değeridir. Geliştirilen bulanık denetleyicide Mamdani çıkarımı ve Toplama yöntemleri kullanılmıştır. Bu iki yönteme ek olarak, bulanık denetleyici ayrıca kullanıcıya ağırlık merkezi, açıortay, maksimumun ortalaması, maksimumun en küçüğü ve maksimumun en büyüğü keskinleştirme yöntemlerinin çıktısını verir. Son olarak, Python programlama dilini ve Tkinter kitaplığını kullanarak, grafik kullanıcı arayüzü, kullanıcının girdilerinin dilsel ifadesini ve üyelik derecesini, nihai bulanık çıktı grafiğini ve tüm durulaştırma yöntemlerinden (GUI) kesin çıktıları görüntüler., The most fundamental characteristic of autonomous vehicles (AVs) is their autonomy. However, due to the dynamic operating environment of the vehicle, their control algorithms may make imprecise, approximate, and unreliable decisions. Therefore, there is a need for the creation of more robust driving algorithms, notably consistent obstacle avoidance algorithms. Occasionally, the vehicle must come to a complete stop in order to avoid obstacles. In this situation, the engine brake control of the car can be engaged. In this study, a fuzzy model was proposed to effectively brake autonomous land vehicles, with an electrical braking system known as rheostatic braking. Since a rheostatic braking system (RBS) is employed, the input values of the fuzzy controller for this designed modeling are vehicle speed and ground slipperiness, and the output value is the rheostat resistance value. In the developed fuzzy controller, Mamdani inference and Aggregation methods were utilized. In addition to these two methods, the fuzzy controller also provides the output of the centroid, bisector, average of the maximum, smallest of the maximum and largest of the maximum sharpening methods to the user. Finally, using the Python programming language and the Tkinter library, the graphical user interface displays the linguistic expression and membership degree of the user's inputs, the final fuzzy output graph, and the exact outputs from all clarification methods (GUI).

Utilizing industrial robots is an efficient method for addressing the labor crisis and advancing industrial technologies. As a result, industrial robots are becoming increasingly popular. Additionally, the widespread use of industrial robots will increase the interest in robot propulsion mechanisms. Legged robots should be primarily investigated because of their potential advantages. Among leg mechanisms, Jansen\"s linkage (JL) has gained popularity due to its organic walking motion, scalable design, and simple drive by rotary input. However, the highly nonlinear nature of JL makes its analysis challenging. The research provides a user-friendly toolbox design that visualizes the toe trajectory and simultaneously calculates the step height by performing a kinematic analysis of the JL using the user-supplied link lengths. In this way, the study contributes significantly to the design phase of legged robots and reduces the amount of time required.

Learning from noisy data is a challenging task especially when the system under consideration has a non-stationary nature. The source of the noise is often assumed to be stationary, however the severity or characteristics of noise may also be time-varying, which causes multiple sources of drift in the collected data. This study introduces a novel adaptive learning rate approach to improve learning when the observations from a non-stationary system is altered by an also non-stationary source of noise. As an example to this approach, we propose Persistence Aware Robust Learner (PeARL), an online learning method that utilizes a novel concept called persistence, which is a local noisiness estimation metric to measure the correspondence of a signal to discrete white noise. Making use of this metric, PeARL is able to adaptively adjust the learning rate for each observation during learning to reduce the effect of noise. With this level of control on the learning rate, noisy instances have less disruptive effect on the maintained estimate. We experimentally evaluate PeARL on (a) systematically generated synthetic data and (b) real-world data, including accelerometer readings collected from people (HASC2010corpus) and current measurements from electric motors collected within the scope of EU-funded research project iRel40. The experiments reveal a favorable region of noise rate, in which the proposed method achieves up to 40% reduction in mean absolute error (MAE).

© 2022 IEEE.One of the applications of Natural Language Processing (NLP) is to process free text data for extracting information. Information extraction has various forms like Named Entity Recognition (NER) for detecting the named entities in the free text. Biomedical named-entity extraction task is about extracting named entities like drugs, diseases, organs, etc. from texts in medical domain. In our study, we improve commonly used models in this domain, such as biLSTM+CRF model, using transformer based language models like BERT and its domain-specific variant BioBERT in the embedding layer. We conduct several experiments on several different benchmark biomedical datasets using a variety of combination of models and embeddings such as BioBERT+biLSTM+CRF, BERT+biLSTM+CRF, Fasttext+biLSTM+CRF, and Graph Convolutional Networks. Our results show a quite visible, 4% to 13%, improvements when baseline biLSTM+CRF model is initialized with pretrained language models such as BERT and especially with domain specific one like BioBERT on several datasets.

© 2022 IEEE.In this study we used machine learning models for predicting individual stock price and volume changes using sentiments from public disclosures and tweets. Public Disclosure Platform (KAP) is the mandated regulatory platform for disclosing news about companies listed in Borsa Istanbul. Investors in Borsa Istanbul use Twitter to express their sentiments about stocks. By combining people\"s sentiment on Twitter and companies\" disclosures, our prediction model predicts the volume and price changes of individual company stocks listed in BIST30. Financial data regarding market conditions consisting of daily price changes of BIST30, DJI, USD, and Gold per Ounce are also added to enhance the prediction accuracy of the model. Our model achieves an maximum of 80% individual stock price prediction accuracy for companies with high social media presence and public disclosure count. We also achieve 74.7% mean volume prediction accuracy across all BIST30 companies.

Financial time series forecasting has been becoming one of the most attractive topics in so many aspects owing to its broad implementation areas and substantial impact. Because of this reason in particular recent decades, various kinds of computational intelligence techniques like convolutional neural networks (CNNs) have been used for financial time series forecasting. However, in experiments reported so far, the number of applications of CNNs for the forecasting of financial time series seems quite a few and also in almost all-studies time sequence effect of time series is not preserved on forecasts because of image transformation. From this point of view, in this paper, by aiming to get better forecasting results and avoiding information loss which may occur the process of image transformation, we suggest a new CNN-based forecasting model that can be applied on some time series and, can successfully extract the features of them in the forecasting process. The proposed CNN forecasting model is composed of three convolutional layers and five full connected layers, also to be able to determine the nonlinear relation between input and output Relu and Elu activation functions have also been used. The suggested framework has been applied to some of the most evaluated financial time series, which are Taiwan Stock Exchange Capitalization Weighted Stock Index (TAIEX) and Financial Time Stock Exchange for London stock market data (FTSE). The results have been evaluated on different aspects as an error criterion, a regression analyses and also a visual demonstration. It has been clearly observed that CNN structure has produced outstanding forecasts compared to some other state-of-the-art forecasting tools such as different kinds of ANN, LSTM, fuzzy-based approaches, and some traditional methods.

© 2022 Elsevier LtdThe Quality Assurance (QA) team verifies software for months before its release decisions. Nevertheless, some crucial bugs remain undetected in manual testing. These bugs would make the system unusable on field, thus merchant loses money then manufacturer loses its customers. Thus, automatic software testing methods have become inevitable to catch more bugs. To locate and repair bugs with an emphasis on the crash scenarios, we present in this work a reinforcement learning (RL) approach for finding and simplifying the input sequence(s) leading to a system crash or blocking, which represents the goal state of the RL problem. We aim at obtaining the shortest input sequence for the same bug so that developers would analyze agent\"s actions causing crashes or freeze. We first simplify the given crash scenario using Recursive Delta Debugging (RDD), then we apply RL algorithms to explore a possibly shorter crashing sequence. We approach the exploration of crash scenarios as a RL problem where the agent first attains the goal state of crash/blocking by executing inputs, then shortens the input sequence with the help of the rewarding mechanism. We apply both model-free on-policy and model-based planning-capable RL agents to our problem. Furthermore, we present a novel RL approach, involving Detected Goal Catalyst (DGC), which reduces the time complexity by avoiding grappling with convergence via stopping learning at a small variance and attaining the shortest crash sequence with an algorithm that recursively removes the unrelated actions. Experiments show DGC significantly improves the learning performance of both SARSA and Prioritized Sweeping algorithms on obtaining the shortest path.

Bilgi ve İletişim Teknolojilerinin (BİT) hızla gelişmesiyle birlikte, çok miktarda kişisel veri oluşmakta, kullanılmakta ve depolanmaktadır. Depolanmakta olan kişisel veriler, son kullanıcıların teknik ve hukuki yönlerden korunmalarını gerektirmektedir. Blokzincir teknolojisi kişisel verilerin gizliliğini korumak ve kontrolünü sağlamak için son yıllarda önemli gelişmeler kaydeden yenilikçi teknoloji olarak görülmektedir. Merkezi olmayan ve Eşler Arası (Peer-to-PeerP2P)2 bir dağıtık dijital defter olan blokzincir teknolojisi, dijital varlıkların tüm işlemlerini depolayabilen, merkezi olmayan, doğrulanabilir ve değiştirilemez bir defter hizmeti sunar. Bir ağdaki katılımcılarla onlara tam olarak güvenmeye gerek kalmadan veri paylaşma konusunda yeni bir yaklaşım sunmaktadır. Yakın zamanda tanıtılan Genel Veri Koruma Yönetmeliği (GDPR) ve Kişisel Verilerin Korunması Kanunu (KVKK), kişisel verilerin nasıl ele alınacağı konusunda büyük değişiklikler getirmektedir. GDPR ve KVKK, kişisel verilerin kullanmasıyla veri denetleyicileri ve işlemciler için yeni koruma zorunlulukları getirmiştir. GDPR ve KVKK, veri koruma mevzuatı kapsamında birliğinin sağlanması için kişisel olarak tanımlanan verilere (PII) daha kolay erişim, silme, düzeltme ve taşıma hakkı verilmesi gibi yeni uygulamalar getirmektedir. GDPR ve KVKK kapsamında, merkezi yapıların hâkim olduğu bir toplumda kişisel veri işleme faaliyetlerinin çoğunlukla merkezi yapılar tarafından gerçekleştirilmesi ve uyulması gereken bir takım usul ve esasları vardır. Ancak blokzincir platformunda ortaya koyulan araştırmalarda kişisel olarak tanımlanan verilerin saklanmasında; merkezi tüzel veya gerçek kişilerin veri saklama, işleme ve silme, gibi uygulamaları gerçekleştirmesi yönünde hazırlanan KVKK ve GDPR hükümlerinin uygulanmasında bazı uyumsuzluklar bulunmaktadır. Bu çalışmada, blokzincirin temel özellikleri detaylandırılmış, kişisel verilerinin kullanımı için blokzincir teknolojisi destekli çözümler açıklanmış ve konuya dair sorunlar ile zorlukları tartışılmıştır. Literatür incelendiğinde; kişisel verilerin günümüzde blokzincir ağında saklanmamasına yönelik tavsiyeler verildiği görülmüştür. Kişisel verilerin KVKK ve GDPR kapsamındaki birincil haklarının, blokzincir teknolojisinin karakteristik yapısına uygun olmadığı, akademik ve uygulamalı araştırmalarda gösterilmiştir. Blokzincir teknolojisindeki gelişme ve güncellemelerin, teknolojinin kendisi ile çelişeceği ve blokzincirin karakteristik özelliğini yok edeceği akademik çevrelerce düşünülmekte ve bundan dolayı temel yapıyı etkilemeyecek (özel anahtarın silinmesi, zincir dışı depolama, karma değeri silinmesi vb.) küçük çaplı değişikliklerin yapılması önerilmektedir. A large amount of personal data is created, used and stored by the rapid development of Information and Communication Technologies (ICT). The personal data being stored requires the end users to be protected from technical and legal aspects. Blockchain technology is seen as an innovative technology that has made significant progress in recent years to protect the privacy and control of personal data. As a decentralized Peer-to-Peer (P2P)3 distributed digital ledger, Blockchain technology provides a decentralized, verifiable and immutable ledger service that can store all transactions of digital assets. It offers a new approach to sharing data with participants in a network without having to fully trust them. The recently introduced General Data Protection Regulation (GDPR) and Personal Data Protection Law (KVKK) bring major changes in how personal data is handled. GDPR and KVKK have brought new protection obligations for data controllers and processors with the use of personal data. In order to ensure unity within the scope of GDPR and KVKK data protection legislation, it brings new applications such as easier access to Personally Identifiable Information (PII), giving the right to delete and rectify and transport. In a society dominated by centralized structures within the scope of GDPR and KVKK, personal data processing activities are mostly carried out by central structures and there are a number of procedures and principles that must be followed. However, there are some inconsistencies in the implementation of the KVKK and GDPR provisions, which are prepared for the central legal or natural persons to carry out applications such as data storage, processing and deletion in the storage of personally-identified data in the researches revealed on the blockchain platform. In this study, the basic features of the blockchain are elaborated, blockchain technology supported solutions for the use of personal data are explained, and the problems and difficulties related to the subject are discussed. When the literature is examined, it has been seen that recommendations are given for not storing personal data in the blockchain network today. It has been shown in academic and applied research that the primary rights of personal data within the scope of KVKK and GDPR are not suitable for the characteristic structure of blockchain technology. Also, it is thought by academic circles that developments and updates in blockchain technology will contradict the technology itself and annihilate the characteristic feature of the blockchain, and therefore it is recommended to make minor changes that not affect the main structure (deleting the private key, off-chain storage, deleting the hash value, etc.).

© 2022 Elsevier LtdFinancial time series prediction problems, for decision-makers, are always crucial as they have a wide range of applications in the public and private sectors. This study presents a cascaded intuitionistic fuzzy model for financial time series prediction. The proposed prediction model has the ability to jointly and simultaneously model linear and nonlinear relationships in financial time series. Thus, it can adapt itself to both linear and non-linear surfaces of the data and can produce satisfactory predictions for financial time series. Moreover, the other reason why to be produced better predictions, the proposed model reckons non-membership degrees in addition to membership degrees in the prediction process. With these aspects, the proposed prediction model is different and superior to all models in the literature. This superiority has been proven by the analysis of 48 different financial time series containing TAIEX, DIJ, SSEC, and IEX data sets. The results have been evaluated in terms of RMSE, MAPE, and MdRAE metrics and some other perspectives as well. The proposed prediction model has achieved progress in prediction performance, up to 80% for TAIEX 2000–2004 datasets, 60% for TAIEX 2008–2018 datasets, approximately 50% for DJI and SSEC, and up to 70% for IEX. All the discussed indicators demonstrated the outstanding prediction performance of the proposed cascaded intuitionistic prediction model compared to some other state-of-the-art prediction tools.

© 2022 IEEE.Health care resources are limited and the efficient utilization of these resources is a must. No show for medical appointments is an everlasting problem that faces health care systems and is a huge waste of resources that can be a hindrance against improving health care services all over the world. This research aims to analyze the factors affecting this problem and create effective predictive models that can help solve it and reduce its social and economical ramifications on health care systems. This study suggests a methodology based on machine learning where different algorithms were utilized and compared and proposes a framework that can be utilized to achieve the best, and most robust classifier in terms of different performance metrics. The objective is to attempt to predict if a patient will attend his/her medical appointment or not.

Internet of Things (IoT) devices have flocked the whole world through the Internet. With increasing mission critical IoT data traffic, attacks on IoT networks have also increased. Many newly crafted attacks on IoT communication require equally intelligent intrusion detection methods to form the first step of countering the attacks. Our work contributes to intrusion detection in IoT networks, by putting state-of-the-art Deep learning methods into service. A BLSTM-GRU Hybrid (BGH) model has been designed to detect eight known IoT network attacks, based on two well-accepted CIC-IDS-2018 and BoT-IoT IoT network traffic datasets. The results of our BGH model in IoT network traffic intrusion detection have been auspicious. The accuracies of prediction on the two datasets are 98.78% and 99.99%. The f1-scores are 98.64% and 99.99%, respectively. The comparison of our results with similar previous studies showed that our BGH model has the best performance ratio (time/accuracy, time/f1-score), where time is the training time of the model. The performance of our proposed model is proof that hybrid Deep Learning methods can prove to be an innovative perspective on Intrusion Detection in IoT networks.

© 2022 IEEE.Natural Language Processing (NLP) is an interdisciplinary field between linguistics and computer science. Its main aim is to process natural (human) language using computer programs. Text classification is one of the main tasks of this field, and they are widely used in many different applications such as spam filtering, sentiment analysis, and document categorization. Nonetheless, there is only very little text classification work in the law domain and even less for the Turkish language. This may be attributed to the complexity within the domain. The length, complexity of documents, and use of extensive technical jargon are some of the reasons that distinguish this domain from others. Similar to the medical domain, understanding these documents requires extensive specialization. Another reason can be the scarcity of publicly available datasets. In this study, we compile sizeable unsupervised and supervised datasets from publicly available sources and experiment with several classification algorithms ranging from traditional classifiers to much more complicated deep learning and transformer-based models along with different text representations. We focus on classifying Court of Cassation decisions for their crime labels. Interestingly, the majority of the models we experiment with could be able to obtain good results. This suggests that although understanding the documents in the legal domain is complicated and requires expertise from humans, it may be relatively easier for machine learning models despite the extensive presence of the technical terms. This seems to be especially the case for transformer-based pre-trained neural language models which can be adapted to the law domain, showing high potential for future real-world applications.

© 2022 IEEE.Deep learning is one of the most significant advances in AI (AI). It is used in a variety of fields due to it has the ability to solve problems that cannot be handled by traditional technologies. The optimization of deep learning relevant to medical images is one of the most important recent advances in image analysis. Several developments have been done on Convolutional Neural Networks to achieve optimal accuracy and increase the learning speed. However, in this paper, we discuss the most recent innovations in convolutional neural networks within Classical method and Quantum method. We briefly provide a snapshot about the architecture, improvements, and principles of both (Classical and Quantum).

© 2022 IEEE.Lately, PM (permanent magnet) motors have started to be used frequently due to advantages such as high efficiency, high power density, high torque, lower losses, lower magnet costs, and more effective control with the development of power electronics. PMSM (permanent magnet synchronous motor) types and BLDCM (brushless direct current motor) types, two members of PM motors, have become frequently used motors in electric vehicles (EVs). In this study, PMSM and BLDCM are compared by optimizing magnet thicknesses, structures, and stator winding turn numbers, to improve the efficiency, using ANSYS Maxwell and ANSYS RMxprt software. In the study, surface mount magnet synchronous motor SPMSM with the inner rotor, V-shaped PM inner rotor IPMSM (interior permanent magnet synchronous motor), BLDCM with inner rotor embedded channel mount magnet, BLDCM with inner rotor embedded magnet, BLDCM with inner rotor surface mount magnet and BLDCM with outer rotor are compared.

© 2022 IEEE.Human gait phase detection has become an emerging field of study due to its impact in various clinical studies. In this study, a system is developed to detect the toe-off, mid-swing, heel-strike, and heel-off phases of a gait cycle in real-time by using a textile-based capacitive strain sensor mounted on the kneepad. Five healthy subjects performed walks including those four phases of the gait at a constant speed and gait distance in a laboratory environment while wearing the kneepad. The phases are labeled according to the gyroscope data of the Inertial Measurement Unit (IMU) located on the kneepad. An Long Short-Term Memory (LSTM) based network is utilized to detect the phases using the capacitance data obtained from the strain sensor. Recognition of four phases with 87 % accuracy is accomplished.

Increasing fossil fuel consumption and consequently the effects of greenhouse gases (GHGs) on the environment and economy are a major concern for all nations and governments. Electric vehicles (EVs) with plug-in capabilities have the potential to ease such problems. However, the extracted power from the grid for charging the EVs\" batteries will significantly impact daily power demand. To satisfy the increasing demand and ensure generation capacity adequacy, the generation expansion planning (GEP) problem is solved to determine the investment decisions for electricity generation sources. Even though there are no centralized utilities for generation planning in most markets, there is still a need to realistically solve the GEP problems and find the optimal investment decisions to tailor the incentives used by most governments to guide the market. There is also a need for a tool to analyze the effect of different charging power levels, charging policies, and penetration levels. The main goal of this paper is to provide a tool to determine realistic optimal investment plans and evaluate different cases. It is also very important to consider the stochastic nature of the electricity demand in GEP problems. We propose a scenario-based stochastic programming model to incorporate the variability in the electricity demand due to EV charging through a set of scenarios generated by Monte Carlo Simulation. The methodology starts with applying a simulation method to generate the electricity demand of EVs by considering all the possible factors affecting EVs\" demand. Each iteration of this simulation represents a possible demand profile as a result of penetrating the EVs into the market. Using all these demand profiles in GEP is preferable, but it is not computationally efficient. Computational tractability is achieved by using the clustering technique to reduce the size of such scenarios. We propose clustering methods to select a representative set from the data sets generated by the simulation and integrate EVs into GEP problems by using the selected set. The GEP models are defined to represent EVs\" demand explicitly and then solved to imply the benefit of the suggested methods. The results show that GEP models with a representative set produce more realistic solutions than the GEP models including only average EVs demand. To select representative sets, different clustering techniques and distance measurements are used and compared with respect to their performances. Two different methods are defined to choose the best number of clusters: the silhouette coefficient method and the elbow method. For each method, five different distance measurement techniques are used. In each of these techniques, three approaches are evaluated for the representative point: Min, Max, and Average. A key contribution of this article is to explore and evaluate the quality of GEP models for each case according to how close the total cost obtained from the GEP model by using clustered load curves to the total cost obtained by using the full data sets generated by simulation.

© 2022 IEEE.The aim of this research is to utilize the particle swarm approach to examine the coordinated design of the unified power flow controller (UPFC) and two power system stabilizers (PSS) in a multi-machine power system. The synchronized proposed control challenge has been presented as an optimal solution with an objective function and constraint equations to complete this goal. Period models in a multi-machine power system were used to assess the effectiveness of the proposed design under distinct operating situations. It has been shown that, even if both PSSs and UPFCs perform well in their own right, power networks stability is adversely affected by an unfavorable impact or negative interaction between controllers if they are not coordinated. Inter-Area oscillations have been efficiently dampened thanks to integrated layout.

© 2022 IEEE.In recent years, smart mobile devices have become indispensable due to the availability of office applications, the Internet, game applications, vehicle guidance or similar most of our daily lives applications in addition to traditional services such as voice calls, SMSs, and multimedia services. Due to Android\"s open source structure and easy development platforms, the number of applications on Google Play, the official Android app store increased day by day. This also brig some security related issues for the end users. The increased popularity of Android operating system on mobile devices, and the associated financial benefits attracted attackers for developing some malware for these devices, which results a significant increase in the number of Android malware applications. To detect this type of security threats, signature based detection (static detection) in generally preferred due to its easy applicability and fast identification ability. Therefore in this study it is aimed to implement an up-to-date, effective, and reliable malware detection system with the help of some deep learning algorithms. In the proposed system, RNN-based LSTM, BiLSTM and GRU algorithms are evaluated on CICInvesAndMal2019 data set which contains 8115 static features for malware detection. Experimental results show that the BiLSTM model outperforms other proposed RNN-based deep learning methods with an accuracy rate of 98.85 %.

—In previous WSN studies, it was observed that there are two common methods used in coverage area calculation: 1) image processing technique or 2) mathematical formulas. calculation of the coverage area, if the sensing ranges of the mobile nodes intersect (overlap) each other, the processing time increases while the calculation result is wrong. In this study, a new method is proposed to calculate the coverage area with relatively high accuracy, including the overlap conditions, but in a shorter time. In this method, the points that were placed as targets to be tracked in the Region of Interest (ROI) in the previous studies are now used in a different way. All covered points by any node are counted to calculate the area. However, in this case, the number of points placed in the area should be much higher than in the previous cases. Coverage performance is calculated by counting the points covered by the WSN nodes and dividing them by the total number of points defined initially. (Coverage rate) First of all, to decide the experiment’s initial conditions, 57 different cases were simulated. Different number of sensor nodes with different radii (25, 50, 100, 200) was considered and then the average of the error rate of all cases was calculated. The lowest error value was obtained in the case of the 2000x2000 area. Therefore, an area of 2000x2000 was preferred in the next steps. It can be concluded that, this method is a promising method to calculate the coverage rate in WSNs. Index Terms—Wireless Sensor Network, sensor deployment, genetic algorithm, evolutionary algorithm, coverage.

© 2022 IEEE.Reinforcement learning methods are mostly constructed on the very assumption that environments are stationary. However, most real world environments are non-stationary; that is, we assume they are composed of several stationary components (i.e., sub-environments or contexts). So, methods with this assumption are not capable of learning non-stationary environments. Reinforcement Learning - Context Detection (RL-CD) method enables the agent to learn the environment without prior information; detect the environment\"s context change points and create a partial model for each context. The underlying environment of this approach is single-agent and has shortcomings for multi-agent learning. In this study, we introduce a new approach called Multi-agent reinforcement learning-context detection (MARL-CD), which can both detect context change points and enable agents to learn non-stationary environments with multi-agent settings. This approach is based on RL-CD approach. MARL-CD is more efficient in terms of detecting context change created by the agents on the environment and detecting the context change of the environment itself. It enables an agent to detect the context changes not only from the change of environment dynamics but also from policy changes of agents in the environment. In the approach in this study, it has been shown by the experimental results that the agents spend 16% less energy and are more efficient than RL-CD in terms of detecting the change points more accurately. —Peki¸stirmeli ögrenme yakla¸sımları, ço ˘ gunlukla or- ˘ tamın duragan olması varsayımıyla etmenin ö ˘ grenmesini konu ˘ alır. Fakat, gerçek hayat uygulamalarında ortam duragan de- ˘ gildir. Birçok dura ˘ gan ortamın bir araya gelmesiyle olu¸san ˘ devingen ortamlardır. Ortamda birden fazla etmen bulunabilir ve bu etmenler de ortamı devingen hale getirmektedir. Peki¸stirmeli ö˘grenme-ba˘glam sezme (RL-CD) [1] yöntemi, etmenin devingen ortam hakkında önsel bir bilgisi olmadan ögrenmesini ˘ ve baglam de ˘ gi¸simlerinin belirlenmesini sa ˘ glayan yakla¸sımdır. ˘ Bu yakla¸sımın temelindeki ortamda tek etmen vardır ve çok etmenli ögrenim için eksiklikleri bulunmaktadır. Bu çalı¸smada ˘ çok etmenli devingen ortamlarda hem baglam de ˘ gi¸sim noktalarını ˘ sezebilen hem de etmenlerin ortamları ögrenebilmesine olanak ˘ saglayan ˘ çok etmenli peki¸stirmeli ö˘grenme-ba˘glam sezme (MARLCD) adında yeni bir yakla¸sım geli¸stirilmi¸stir. Bu yakla¸sım RLCD yöntemini temel alır. Çok etmenli ögrenmede, etmenlerin ˘ ortam üzerinde olu¸sturdukları devingenligi sezmesi ve ba ˘ glam ˘ degi¸sikli ˘ gini belirlemesi yönüyle daha verimlidir. Ba ˘ glamdaki ˘ degi¸siklikleri yalnızca ortam dinamiklerinin de ˘ gi¸siminin yanı sıra ˘ ortamdaki etmenlerin politika degi¸siklikleriyle de belirleyebilme- ˘ sini saglar. Bu çalı¸smadaki yakla¸sımda, etmenler enerjilerini %16 ˘ daha az harcayarak ve degi¸sim noktalarını daha do ˘ gru sezmesi ˘ açısından RL-CD’ye daha verimli oldugu, deney sonuçları ile ˘ gösterilmi¸stir

© 2022 IEEE.It is very important to fight the Mediterranean fruit fly pest in orchards in our country. If this fight is not given, serious economic losses occur. Traps are currently being set up on trees to detect pests. It is important that these traps are checked by farmers at regular intervals in terms of pesticide applications of fruits. This process increases the workload of farmers. Therefore, a smart trap design that automatically detects the number of Mediterranean fruit flies is proposed in this study. The smart trap is implemented as an embedded system using the STM 32F746GDISCOVERY development board with an Arm Cortex M7 processor on it. Via the proposed system, quantitative data can be obtained about the population of Mediterranean fruit flies in the determined region. Machine learning is performed on the proposed embedded system. Therefore, a deep learning-based structure that can work on the embedded system has been designed. Obtained preliminary results indicate that the proposed system can be used for Mediterranean fruit fly detection and automatic determination of its number in a given region. Ülkemizde meyve bahçelerinde Akdeniz meyve sinegi˘ zararlısı ile mücadele çok önemlidir. Bu mücadelenin verilmemesi durumunda ciddi ekonomik kayıplar olu¸smaktadır. ¸Su anda zararlının tespit edilebilmesi için meyve agaçlarının üzerine tu- ˘ zaklar kurulmaktadır. Bu tuzakların düzenli aralıklarla çiftçiler tarafından kontrol edilmesi, meyvelerin ilaçlanma uygulamaları açısından önem arz etmektedir. Bu i¸slem çiftçilerin i¸s yükünü arttırmaktadır. Bu çalı¸smada, Akdeniz meyve sinegi zararlısı ˘ sayısını otomatik olarak tespit eden bir akıllı tuzak tasarımı önerilmi¸stir. Akıllı tuzak, Arm Cortex M7 i¸slemcisi barındıran STM 32F746GDISCOVERY geli¸stirme kartı kullanılarak gömülü sistem olarak gerçeklenmi¸stir. Önerilen sistem sayesinde belirlenen bölgede Akdeniz meyve sinegi zararlısının popülasyonu ˘ hakkında nicel veriler elde edilebilmektedir. Önerilen gömülü sistem üzerinde makine ögrenmesi yapılmaktadır. Bunun için gö- ˘ mülü sistem üzerinde çalı¸sabilecek derin ögrenme temelli bir yapı ˘ kurgulanmı¸stır. Elde edilen ön sonuçlar önerilen sistemin Akdeniz meyve sinegi tespiti ve belirli bir alandaki sayısının otomatik ˘ olarak bulunması için kullanılabilecegini göstermektedir.

© 2022 IEEE.Deep Learning (DL) is the most efficient technique to handle a wide range of challenging problems such as data analytics, diagnosing diseases, detecting anomalies, etc. The development of DL has raised some privacy, justice, and national security issues. Deepfake is a DL-based application that has been very popular in recent years and is one of the reasons for these problems. Deepfake technology can create fake images and videos that are difficult for humans to recognize as real or not. Therefore, it needs to be proposed some automated methods for devices to detect and evaluate threats. In another word, digital and visual media must maintain their integrity. A set of rules used for Deepfake and some methods to detect the content created by Deepfake have been proposed in the literature. This paper summarizes what we have in the critical discussion about the problems, opportunities, and prospects of Deepfake technology. We aim for this work to be an alternative guide to getting knowledge of Deepfake detection methods. First, we cover Deepfake history and Deepfake techniques. Then, we present how a better and more robust Deepfake detection method can be designed to deal with fake content.

© 2022 IEEE.Using convolutional neural networks, this thesis aims to create a system for fully automated identification of bird species based on spectrogram images. Spectrogram analysis is more difficult when trying to make an advance identification of a bird species. On a publicly available dataset of 8000 audio examples, we\"ve begun by analyzing the challenges of bird species detection, segmentation, and classification to achieve our goal. It has been determined also that deep learning-based technique CNN with Fully convolutional learning calls for easier results because it eliminates the possible future modelling error caused by an imprecise knowledge of bird species and works well on coding in cohesion with the spectral analysis kernel using the librosa library. We have concluded. After obtaining the dataset from the open-source repository, it is then processed locally. For training, testing, and validation we used a subset of the dataset of 8000 sound samples. We offered a method relying on a CNN reset learned that proved to be very quick and optimum because it was first needing the spectrogram analytic kernel to learn what to class in bird species, and then it gets the system trained on features extracted. In a novel 9-step implementation, a bird species spectrogram can be detected from an audio sample. There was a loss of less than 0.0063, and the conditioning workouts accuracy is 0.9895 for the system, 0.9 as precision, and training and validation use 50 epochs in system.

Teknolojinin gelişmesi ve internet kullanıcı sayısındaki artışla orantılı olarak siber suçlarda da artış gözlemlenmiştir. Birçok farklı siber saldırı tekniği bulunmaktadır. Bu saldırı tekniklerinden biri olan kötü amaçlı web siteleri, siber saldırılar ve dolandırıcılık olaylarında önemli rol oynamaktadır. İnternette masum görünen bir bağlantıya tıklamak veya e-posta ve mesaj yoluyla gönderilen bir web sayfasını ziyaret etmek arka planda kimlik avı kampanyalarının başlatılmasına, kötü amaçlı yazılımların, casus yazılımların, fidye yazılımların indirilmesine ve ciddi parasal kayıplar oluşmasına yol açar. Dolayısıyla bu tehditlerin etkin bir şekilde tespit edilmesi ve önlenmesi bireyler, kurumlar ve hükümetler için oldukça önemli bir konu haline gelmiştir. Kara listeye dayalı yöntemler, kötü amaçlı URL\"leri tanımlamak için kullanılan standart yöntemlerden biridir. Ancak kara listeler hiçbir zaman kapsamlı değildir ve yeni oluşturulan URL\"leri algılama yeteneğinden yoksundur. Kara listeye dayalı yöntemlerin mevcut ihtiyacı ve eksiklikleri de göz önünde bulundurularak bu çalışmada toplulukla öğrenme yöntemleri kullanılarak bir sınıflandırma yaklaşımı önerilmiştir. Çalışmada iyi huylu ve kötü huylu URL’lerden elde edilmiş 79 sözcüksel özellik içeren Kanada Siber Güvenlik Enstitüsü\"nün URL veriseti (ISCX-URL-2016) üzerinde çalışılmıştır. Verisetinde benign, spam, phishing, malware ve defacement olmak üzere beş farklı URL türü bulunmaktadır. Toplam 7781 iyi huylu ve 28.917 tane zararlı URL kaydı üzerinde zararlı, zararsız etiketleri kullanılarak ikili sınıflandırma işlemi ve beş farklı etiket bilgisi kullanılarak çoklu sınıflandırma işlemi gerçekleştirilmiştir. Makine öğrenmesi yöntemlerinden Rastgele Orman algoritması uygulanan yöntemin başarısının sınanması için 10-katlamalı çapraz doğrulama (10-fold cross validation) ile birlikte kullanılmıştır ve 10 temel bileşen kullanılarak ikili sınıflandırma problemi için ortalama %99.42, çoklu sınıflandırma problemi için ortalama %95.68 doğruluk değeri elde edilmiştir. Böylece her gün yeni web sitelerinin katıldığı bu dinamik internet ağını kötü niyetli tasarlanmış web sitelerinden korumaya yönelik yüksek başarım oranına sahip bir model önerisi sunulmuştur. In parallel with the development of technology and the increase in the number of internet users, an increase in cybercrime has been observed. There are many different cyberattack techniques. Malicious websites, one of these attack techniques, play an important role in cyberattacks and fraud events. Clicking on an innocent-looking link on the Internet or visiting a web page sent via email or text will result in phishing campaigns being launched in the background, downloading malware, spyware, ransomware, and serious monetary losses. Therefore, effective detection and prevention of these threats has become a very important issue for individuals, institutions and governments. Blacklist-based methods are one of the standard methods used to identify malicious URLs. However, blacklists are never comprehensive and lack the ability to detect newly created URLs. Considering the current needs and deficiencies of blacklist-based methods, a machine learning based classification approach was used in this study to combat malicious URLs. In the study, the URL data set of the Canadian Cyber Security Institute (ISCX-URL-2016) was studied, which contains 79 lexical features obtained from benign and malignant URLs. There are five different URL types in the dataset: benign, spam, phishing, malware and defacement. A binary classification process using harmless, malicious labels and a multi-classification process using five different labels information was performed on a total of 7781 benign, harmless and 28,917 malicious URL records. Random Forest algorithm, one of the machine learning methods, used together with 10-fold cross validation to validate the success of the applied method, and an average accuracy value of 99.42% for the binary classification problem and 95.68% for the multiple classification problem was obtained. Thus, a model proposal with a high-performance rate is presented to protect this dynamic internet network, where new websites are added every day, from maliciously designed websites.

© 2021 IEEE.Coronavirus usually transmits from the animal to the human, but now, the virus transmission is between persons. Therefore, scientists and researchers are trying to develop several types of machine learning methods to defend against COVID-19. Medical images play a significant role in this time due to they can be used to recognize COVID-19 accurately. However, in this paper, we used X-Ray images, the images undergone to sharpening techniques to increase the results further. The texture techniques named local binary pattern (LBP) have been used in order to extract features. The features obtained were applied to the support vector machine (SVM). The results we achieved were 100% for all performance measurements. Our results were conspicuously superior compared to the state-of-the-art papers published.

Dynamic systems are highly complex and hard to deal with due to their subject-and time-varying na-ture. The fact that most of the real world systems/events are of dynamic character makes modeling and analysis of such systems inevitable and charmingly useful. One promising estimation method that is ca-pable of unlearning past information to deal with non-stationarity is Stochastic Learning Weak Estimator (SLWE) by Oommen and Rueda (2006). However, due to using a constant learning rate, it faces a trade-off between plasticity and stability. In this paper, we model SLWE as a random walk and provide rigorous theoretical analysis of asymptotic behavior of estimates to obtain a statistical model. Utilizing this model, we detect changes in stationarity to switch between exploratory and exploitative learning modes. Exper-imental evaluations on both synthetic and real world data show that the proposed method outperforms related algorithms in different types of drifts. (c) 2022 Elsevier Ltd. All rights reserved.

Collision risk assessment and collision avoidance of vessels have always been an important topic in ocean engineering. Decision support systems are increasingly becoming the focus of many studies in the maritime industry today as vessel accidents are often caused by human error. This study proposes an anticollision decision support system that can determine surrounding obstacles by using the information received from radar systems, obtain the position and speed of obstacles within a certain time period, and suggest possible routes to prevent collisions. In this study we use a neural network to predict the subsequent positions of surrounding vessels, a fuzzy logic system to obtain the risk of collision, and a particle swarm optimization algorithm to find the safe and shortest path for collision avoidance.

Online anomaly detection and identification is a major task of many Industry 4.0 applications. Electric motors, being one of the most crucial parts of many products, are subjected to end-of-line tests to pick up faulty ones before being mounted to other devices. With this study, we propose a Syntactic Pattern Recognition based approach to online anomaly detection and identification on electric motors. Utilizing Variable Order Markov Models and Probabilistic Suffix Trees, we apply both unsupervised and supervised approaches to cluster motor conditions and diagnose them. Besides being explainable, the diagnosis method we propose is completely online and suitable for parallel computing, which makes it a favorable method to use synchronously with a physical test system. We evaluate the proposed method on a benchmark dataset and on a case study, which is being worked on within the scope of a European Union funded research project on reliability.