Yazılım Mühendisliği Bölümü Koleksiyonu
Bu koleksiyon için kalıcı URI
Güncel Gönderiler
Öğe Textual Authenticity in the AI Era: Evaluating BERT and RoBERTa with Logistic Regression and Neural Networks for Text Classification(Institute of Electrical and Electronics Engineers Inc., 2024) Hazim, Layth Rafea; Ata, OğuzAI-generated content impersonating human writing is an issue that has gained attention as AI spreads its wings. This particular study serves as a comparison between the existing Logistic Regression and Feedforward Neural Networks (FNNs) by employing sentence-BERT-appended models and RoBERTa in the determination of authenticity in a given text. A dataset that was balanced between human-written and AI-generated texts was utilized. Techniques like tokenization and normalization were first used, followed by feature extraction using transformer-based models. Cross-validation and confusion matrix analysis that used measures such as accuracy, precision, recall, F1 score, and ROC AUC were included to guarantee the models' robustness. The hybrid RoBERTa-FNN model that was deposed challengers was the most outstanding model in respect of precision and recall, and the highest accuracy (99.95%) was obtained as mentioned in the data. The improved performance serves as a proof of how effectively RoBERTa uses its embeddings to represent context on the fine-grained level required for this kind of text classification. This work is a stepping stone to the creation of strong AI text detection systems, besides our advancements to the knowledge of models and embedding performance with respect to text classification. The results lay emphasis on the selection of model configuration and the embedding technique, as they are the key factors in achieving the best results in practical applications. © 2024 IEEE.Öğe AI-Driven Triage: A Graph Neural Network Approach for Prehospital Emergency Triage Patients in IoMT-Based Telemedicine Systems(Institute of Electrical and Electronics Engineers Inc., 2024) Jasim, Abdulrahman Ahmed; Ata, Oguz; Salman, Omar HusseinImproper triage, overcrowding in the emergency department (ED), and long wait times for patients are some of the problems that arise from emergency triage due to a lack of medical and human resources. In emergency departments, proper patient triage is crucial for determining the urgency and appropriateness of treatment by evaluating the severity of each patient's condition. In most cases, a human operator will use the knowledge and data gathered from patient management to perform triage tactics. Hence, it is a method with the possibility of causing problems in high-priority interconnections. Currently, typical triage systems primarily rely on human recommendations, which are susceptible to personal convictions and error. Methods that maximize data collection and remove mistakes in patient triage processing are being developed with an increasing amount of interest in utilizing artificial intelligence (AI). We develop and deploy an AI module to control the distribution of emergency triage patients in EDs. We use data from past emergency room visits to educate the medical decision-making process. Patients are properly sorted into triage groups because of data including vital signs, symptoms, and an ECG. Test results have shown that the proposed algorithm is better than the currently used state-of-the-art triage techniques. In our opinion, the methods used will help healthcare professionals forecast the severity index, which in turn will help them direct the proper patient management protocols and allocate resources. © 2024 IEEE.Öğe DFT study on the his-tag binding affinity of metal Ions in modeled hexacationic metal complexes(2024) Yıldız, Alparslan Numan; Kasapbaşı, E. Esra; Yurtsever, Zeynep; Yurtsever, MineHistidine oligomers (His-tags) are commonly used as affinity tags in recombinant protein purification to enable in vitro experimental studies, including biochemical and biophysical assays and structure determination. His-tags enable protein purification by specifically and efficiently coordinating bivalent metal ions present in the purification resins, such as Cu2+, Zn2+, and Ni2+. Although His-tags, combined with Ni2+-based resins, are widely used due to their biophysical properties and commercial availability, the structure and nature of the metal cation coordination have remained unclear. In this study, the chemical structure of metal-coordinating His-tags was modeled to elucidate the metal preferences for better binding and to determine the structural changes that occur upon metal coordination. 6His-tag is a string of 6 histidine residues usually coordinated to bivalent metal ions (M2+), such as Ni2+, Zn2+, and Cu2+, through the nitrogen atoms of the imidazole rings. The metals complete their octahedral coordination shell with the lone pair of electrons of the oxygen atoms of the resin to which they are attached to. The resin was modeled as an oxalaldehyde group with the closed formula of (OCH)4. The geometry optimizations were carried out using the DFT method at the B3LYP/6-31g (d, p) level and in implicit water using the IEFPCM solvent model. The impact of Ni2+, Zn2+, and Cu2+ metals on the resin binding ability of 6His-tags was tested by adding amino acids to the N-terminus of metal-coordinated hexahistidine chains. The activity of the metals as electron donors or acceptors to the coordination bonds that they form was estimated by calculating the NBO energies. The complexation enthalpies of metal-bearing resin with hexahistidine, naked or having an amino acid tail, were calculated. Our results showed that Ni2+ has the highest affinity for the recombinant 6His-tag.Öğe The web applications cross site scripting attacks and preventions using machine learning technique(2024) Alyasin, Eman Ibrahim; Ata, Oğuz; Özturk, Bilal A.Web applications are utilized everywhere these days to share services and data online. Because companies deal with sensitive data, hackers have found them attractive targets. Vulnerabilities persist despite the numerous security procedures we've created to safeguard these applications. Major security issues have been identified in web applications used by various organizations, such as banks, healthcare providers, finance companies, and retail businesses. Cross-site scripting (XSS) attacks are one of the most significant issues, according to a report from White Hat Security. These attacks enable hackers to execute harmful programs on a user's web browser, resulting in issues such as the theft of data, cookies, passwords, and credit card numbers. This study focuses on the primary weaknesses present in contemporary web applications, particularly XSS attacks. We go over the many kinds of XSS attacks, provide instances from the real world, and describe how they operate. We also examine defenses against these attacks, discussing what works and what doesn't.Öğe An in vitro assessment of ionizing radiation impact on the efficacy of radiotherapy for breast cancer(2024) Girgin, Merve; Akat, Ayberk; Akgül, Büşra; Nalbant, Nilgül; Karaçetin, Didem; Abamor, Emrah Şefik; Uğurel, Osman Mutluhan; Turgut-Balık, DilekObjectives Ionizing radiation is still one of the most effective treatment options for various cancers. It is possible to reduce the side effects of this effective treatment method and increase the chance of success by elucidating the responses it creates at the molecular level in the cell. This study aims to investigate of the molecular effects of therapeutic ionizing radiation on breast cancer, which is the most prevalent cancer type. Methods MDA-MB-231 and MCF7 cell lines were irradiated with 4 and 8 Gy ionizing radiation and monitored for up to 7 days. RNA was collected at 48 and 96 h, when cellular molecular mechanisms became most evident, and quantitative expression levels of microRNAs (miR-208a, miR-124, miR-145), for which cancer-radiation associations have been determined from existing literature and databases, were evaluated. Results Exposure to ionizing radiation resulted in a dose-dependent reduction in cell viability in both MCF7 and MDA-MB-231 breast cancer cell lines. Furthermore, microRNA expression analysis revealed notable changes at all levels. The research demonstrates that miR-208a, miR-145, and miR-124 are crucial in the biological response to ionizing radiation. Conclusions Therapeutic ionizing radiation profoundly affects cell viability and microRNA expression in breast cancer cell lines, showing dose and time-dependent effects. The observed microRNA expression patterns suggest potential biomarkers for radiation response and therapeutic targets to improve radiotherapy efficacy. Further in vivo validation and exploration of these microRNAs' roles in modulating cellular response to ionizing radiation are needed.Öğe Performance model for factory automation in 5G networks(2022) Wang, Jiao; Weitzen, Jay; Bayat, Oğuz; Sevindik, Volkan; Li, MingzheThe fifth generation (5G) of mobile networks is emerging as a key enabler of modern factory automation (FA) applications that ensure timely and reliable data exchange between network components. Network slicing (NS), which shares an underlying infrastructure with different applications and ensures application isolation, is the key 5G technology to support the diverse quality of service requirements of modern FA applications. In this article, an end-to-end (E2E) NS solution is proposed for FA applications in a 5G network. Regression approaches are used to construct a performance model for each slice to map the service level agreement to the network attributes. Interference coordination approaches for switched beam systems are proposed to optimize radio access network (RAN) performance models. A case study of a non-public network is used to show the proposed NS solution. Simulation result shows that for services with different QoS requirements, different IC approaches should be used as optimization methods. Design prediction using regression approach has been evaluated and shows that the prediction successful rate increases when more existing data are used.Öğe Diagnosing cervical cancer using machine learning methods(Institute of Electrical and Electronics Engineers Inc., 2022) Coşar Soğukkuyu, Derya Yeliz; Ata, OğuzPre-diagnosis of any kind of cancer type has critical impact on person's life. According to World Health Organization Cervical cancer is still one of the most common gynecologic cancers in the world that affects women life. if medical experts focus on early diagnosis of the disease which is detecting symptomatic patients as early as possible, patients will have the best chance for successful treatment since Cervical Cancer is preventable. When cancer treatment is delayed, mortality rate decreases, and treatment becomes more complicated and expensive. Currently, systems based on Artificial Intelligence are used for decision making Machine learning techniques let automated detection of cervical cancer run more quickly and efficiently. In this study a novel ensemble approach is presented to predict the risk of cervical cancer by developing hybrid machine learning model. Multiple performance measurements such as Accuracy, precision score, recall score, and F1 are performed to evaluate the novel model. The results indicate that the proposed novel model can be effectively used to pre diagnosis of Cervical cancer with accuracy 97%.Öğe Real-time data of COVID-19 detection with IoT sensor tracking using artificial neural network(Computers and Electrical Engineering, 2022) Ata, Oğuz; Mohammedqasem, Roa'a; Mohammedqasim, HayderThe coronavirus pandemic has affected people all over the world and posed a great challenge to international health systems. To aid early detection of coronavirus disease-2019 (COVID-19), this study proposes a real-time detection system based on the Internet of Things framework. The system collects real-time data from users to determine potential coronavirus cases, analyses treatment responses for people who have been treated, and accurately collects and analyses the datasets. Artificial intelligence-based algorithms are an alternative decision-making solution to extract valuable information from clinical data. This study develops a deep learning optimisation system that can work with imbalanced datasets to improve the classification of patients. A synthetic minority oversampling technique is applied to solve the problem of imbalance, and a recursive feature elimination algorithm is used to determine the most effective features. After data balance and extraction of features, the data are split into training and testing sets for validating all models. The experimental predictive results indicate good stability and compatibility of the models with the data, providing maximum accuracy of 98% and precision of 97%. Finally, the developed models are demonstrated to handle data bias and achieve high classification accuracy for patients with COVID-19. The findings of this study may be useful for healthcare organisations to properly prioritise assets.Öğe Weather routing optimization system for the purpose of reducing fuel consumption and navigation time of merchant ships(Politeknik Dergisi, 2021) Baba, Ahmet FevziAim To build a decision support system to find the optimal routes in different sea and weather conditions. Design & Methodology In this study, the Aegean Sea is selected as environment. For calculations, CSGA Algorithm is used. Originality This study is the first study which uses CSGA Algorithm fort he purpose of weather routing. Findings The system was able to find optimal routes in rough weather conditions. Conclusion A weather routing system which obtains sea and weather data from Inmarsat and Copernicus services and finds optimal routes using CSGA is proposed. Declaration of Ethical Standards The author(s) of this article declare that the materials and methods used in this study do not require ethical committee permission and/or legal-special permission.Öğe Comment on "synthesis, characterization and growth mechanism of flower-like vanadium carbide hierarchical nanocrystals"(2012) Presser, Volker; Vakifahmetoglu, CekdarThis Letter is in response to a recent paper by Ma et al. (CrystEngComm, 2010, 12, 750-754) which arguably studied vanadium carbide nanostructures whereas all available evidence indicates the study of vanadium oxide. We feel that it is important to communicate to the community several inconsistencies so that the interesting material reported can be seen in the right light, especially with several groups nowadays having reported similar structures from vanadium oxide synthesis. © 2012 The Royal Society of Chemistry.Öğe Effects of hurricanes Irene and Sandy in New Jersey: Traffic patterns and highway disruptions during evacuations(Kluwer Academic Publishers, 2015) Li, Jian; Özbay, Kaan; Bartın, BekirThis paper describes a quantitative analysis of traffic patterns and highway disruptions during Hurricane Irene and Sandy evacuations in New Jersey (NJ). This empirical study is based on multiple traffic and event data collected by various transportation agencies in NJ. In the first part of the paper, the temporal and spatial traffic patterns in NJ during Irene and Sandy evacuations were explored, and a comparative assessment of evacuation departure models was conducted based on the empirical traffic data. In the second part, we explored the frequency and geographic distribution of highway disruptions (vehicle accidents/incidents, incidents such as downed trees or road flooding caused by extreme winds and heavy rains, and highway bottlenecks) during Irene and Sandy evacuations and pre-landfall periods. The empirical patterns observed in this study can be used to improve real-world emergency response operations and evacuation models. The empirical findings may also benefit hurricane evacuation planning in areas with similar circumstances as NJ. © 2015, Springer Science+Business Media Dordrecht.Öğe Multi-level image thresholding based on social spider algorithm for global optimization(Springer Science+Business Media B.V., 2019) Rahkar, Farshi, Taymaz; Orujpour, MohannaThresholding is one of the simplest and popular technique for segmenting images. Maximum between-class variance (Otsu’s) method is one of the well-known and widely used method in case of segmentation. Not only Otsu could be used for bi-level thresholding but also it could be extended to multi-level image thresholding. Finding the optimum threshold values in multi-level case is very time consuming process, thus optimization algorithm can deal with this problem. In this paper social spider algorithm for global optimization has been used for maximizing the between-class variance to carry out multi-level image thresholding. Experimental outcomes have demonstrated that the proposed method is capable of estimating threshold values and yield satisfying outcome. © 2019, Bharati Vidyapeeth's Institute of Computer Applications and Management.Öğe Improving stock prediction accuracy using CNN and LSTM(Institute of Electrical and Electronics Engineers Inc., 2020) Rasheed, Jawad; Jamil, Akhtar; Hameed, Alaa Ali; Ilyas, Muhammad; Özyavaş, Adem; Ajlouni, NaimStock price modeling and prediction is a challenging task due to its non-stationary and dynamic nature of data. Developing an accurate stock prediction method can help investors in making profitable decisions by reducing the investment risks. This paper proposes a deep learning-based method for significantly improving the stock prediction accuracy using deep learning-based methods. Two well-known methods were investigated, namely one dimensional Convolutional Neural Network (1D-CNN) and the Long Short-Term Memory (LSTM). In addition, we also investigated the effect of dimensionality reduction using principal component analysis (PCA) on the prediction accuracy of both 1D-CNN and LSTM. Two separate experiments were performed for each method, one with PCA and one without PCA. The experimental results indicated that LSTM with PCA produced the best results with mean absolute error (MAE) of 0.032, 0.084, and 0.044 while a root mean square error (RMSE) of 0.0643, 0.172, 0.079 on Apple Inc., Amerisource Bergen Corporation, and Cardinal Health datasets. The LSTM network with PCA took an average of 421.8s for training. Contrarily, 1D-CNN model with PCA performed better in terms of computational time as it took only 37s for training and attained MAE of 0.039 and RMSE of 0.0706 on Apple Inc. dataset. Similarly, 1D-CNN took 36.5s for training while achieving 0.099 MAR and 0.2021 RMSE on Amerisource Bergen Corporation dataset, while 37.5s for training that secured 0.067 MAE and 0.1037 RMSE on Cardinal Health dataset. © 2020 IEEE.Öğe Broadband coherent perfect absorber with PT-symmetric 2D-materials(Academic Press Inc Elsevier Science, 2019) Sarısaman, Mustafa; Taş, MuratWe suggest graphene and a two-dimensional (2D) Weyl semimetal (WSM) as 2D materials for the realization of a broadband coherent perfect absorber (CPA) respecting overall PT-symmetry. We also demonstrate the conditions for mutually equal amplitudes and phases of the left and right incoming waves to realize a CPA. 2D materials in our system play the role to enhance the absorption rate of a CPA once the appropriate parameters are inserted in the system. We show that a 2D WSM is more effective than graphene in obtaining the optimal conditions. We display the behavior of each parameter governing the optical system and show that optimal conditions of these parameters give rise to enhancement and possible experimental realization of a broadband CPA-laser. (C) 2019 Elsevier Inc. All rights reserved.Öğe Interactive lane closure and traffic information tool based on a geographic information system(Natl Acad Sciences, 2012) Bartın, Bekir; Özbay, Kaan; Mudigonda, SandeepThis paper describes the development of the Rutgers Interactive Lane Closure Application (RILCA), an interactive computer tool to plan lane closures for work zones. This tool provides traffic engineers with a computerized and easy-to-use lane closure application, along with other useful features. RILCA was developed with the Arc View geographic information system (GIS) software package as the main development environment. Arc View displays the interactive GIS map of the New Jersey Turnpike, Garden State Parkway, and other major freeways in New Jersey and its surrounding network. RILCA provides various analysis and visualization options to plan lane closures interactively, obtain traffic volume information, and conduct accident queries. RILCA was tested successfully, and New Jersey Department of Transportation (DOT) and New Jersey Turnpike Authority engineers now use it with all of the features presented in this paper. RILCA received an annual implementation award from the New Jersey DOT in 2009. The award recognized RILCA's usefulness and potential to improve the efficiency of traffic operations as a practical decision-support tool for lane closure decisions.Öğe Empirical evacuation response curve during hurricane Irene in Cape may county, New Jersey(Sage Publications Inc, 2013) Li, Jian; Özbay, Kaan; Bartın, Bekir; Iyer, Shrisan; Carnegie, Jon A.Understanding evacuation response behavior is critical for public officials in deciding when to issue emergency evacuation orders for an impending hurricane. Such behavior is typically measured by an evacuation response curve that represents the proportion of total evacuation demand over time. This study analyzes evacuation behavior and constructs an evacuation response curve on the basis of traffic data collected during Hurricane Irene in 2011 in Cape May County, New Jersey. The evacuation response curve follows a general S-shape with sharp upward changes in slope after the issuance of mandatory evacuation notices. These changes in slope represent quick response behavior, which may be caused in part by an easily mobilized tourist population, lack of hurricane evacuation experience, or the nature of the location, in this case a rural area with limited evacuation routes. Moreover, the widely used S-curves with different mathematical functions and the state-of-the-art behavior models are calibrated and compared with empirical data. The results show that the calibrated S-curves with logit and Rayleigh functions fit empirical data better. The evacuation behavior analysis and calibrated evacuation response models from this hurricane evacuation event may benefit evacuation planning in similar areas. In addition, traffic data used in this study may also be valuable for the comparative analysis of traffic patterns between the evacuation periods and regular weekdays and weekends.Öğe Comparative analysis of different distributions dataset by using data mining techniques on credit card fraud detection(Univ Osijek, Tech Fac, 2020) Ata, Oğuz; Hazim, LaythBanks suffer multimillion-dollars losses each year for several reasons, the most important of which is due to credit card fraud. The issue is how to cope with the challenges we face with this kind of fraud. Skewed "class imbalance" is a very important challenge that faces this kind of fraud. Therefore, in this study, we explore four data mining techniques, namely naive Bayesian (NB),Support Vector Machine (SVM), K-Nearest Neighbor (KNN) and Random Forest (RF), on actual credit card transactions from European cardholders. This paper offers four major contributions. First, we used under-sampling to balance the dataset because of the high imbalance class, implying skewed distribution. Second, we applied NB, SVM, KNN, and RF to under-sampled class to classify the transactions into fraudulent and genuine followed by testing the performance measures using a confusion matrix and comparing them. Third, we adopted cross-validation (CV) with 10 folds to test the accuracy of the four models with a standard deviation followed by comparing the results for all our models. Next, we examined these models against the entire dataset (skewed) using the confusion matrix and AUC (Area Under the ROC Curve) ranking measure to conclude the final results to determine which would be the best model for us to use with a particular type of fraud. The results showing the best accuracy for the NB, SVM, KNN and RF classifiers are 97,80%; 97,46%; 98,16% and 98,23%, respectively. The comparative results have been done by using four-division datasets (75:25), (90:10), (66:34) and (80:20) displayed that the RF performs better than NB, SVM, and KNN, and the results when utilizing our proposed models on the entire dataset (skewed), achieved preferable outcomes to the under-sampled dataset.Öğe Multi-modal forest optimization algorithm(Springer London Ltd, 2020) Orujpour, Mohanna; Feizi-Derakhshi, Mohammad-Reza; Rahkar-Farshi, TaymazMulti-modal optimization algorithms are one of the most challenging issues in the field of optimization. Most real-world problems have more than one solution; therefore, the potential role of multi-modal optimization algorithms is rather significant. Multi-modal problems consider several global and local optima. Therefore, during the search process, most of the points should be detected by the algorithm. The forest optimization algorithm has been recently introduced as a new evolutionary algorithm with the capability of solving unimodal problems. This paper presents the multi-modal forest optimization algorithm (MMFOA), which is constructed by applying a clustering technique, based on niching methods, to the unimodal forest optimization algorithm. The MMFOA operates by dividing the population of the forest into subpopulations to locate existing local and global optima. Subpopulations are generated by the Basic Sequential Algorithmic Scheme with a radius neighborhood. As population size is self-adaptive in MMFOA, population size can be increased in functions with too many local and global optima. The proposed algorithm is evaluated by a set of multi-modal benchmark functions. The experiment results show that not only is the population size low, but also that the convergence speed is high, and that the algorithm is efficient in solving multi-modal problems.Öğe Battle royale optimization algorithm(Springer London Ltd, 2021) Farshi, Taymaz RahkarRecently, several metaheuristic optimization approaches have been developed for solving many complex problems in various areas. Most of these optimization algorithms are inspired by nature or the social behavior of some animals. However, there is no optimization algorithm which has been inspired by a game. In this paper, a novel metaheuristic optimization algorithm, named BRO (battle royale optimization), is proposed. The proposed method is inspired by a genre of digital games knowns as "battle royale." BRO is a population-based algorithm in which each individual is represented by a soldier/player that would like to move toward the safest (best) place and ultimately survive. The proposed scheme has been compared with the well-known PSO algorithm and six recent proposed optimization algorithms on nineteen benchmark optimization functions. Moreover, to evaluate the performance of the proposed algorithm on real-world engineering problems, the inverse kinematics problem of the 6-DOF PUMA 560 robot arm is considered. The experimental results show that, according to both convergence and accuracy, the proposed algorithm is an efficient method and provides promising and competitive results.Öğe A multimodal particle swarm optimization-based approach for image segmentation(Pergamon-Elsevier Science Ltd, 2020) Farshi, Taymaz Rahkar; Drake, John H.; Özcan, EnderColor image segmentation is a fundamental challenge in the field of image analysis and pattern recognition. In this paper, a novel automated pixel clustering and color image segmentation algorithm is presented. The proposed method operates in three successive stages. In the first stage, a three-dimensional histogram of pixel colors based on the RGB model is smoothened using a Gaussian filter. This process helps to eliminate unreliable and non-dominating peaks that are too close to one another in the histogram. In the next stage, the peaks representing different clusters in the histogram are identified using a multimodal particle swarm optimization algorithm. Finally, pixels are assigned to the most appropriate cluster based on Euclidean distance. Determining the number of clusters to be used is often a manual process left for a user and represents a challenge for various segmentation algorithms. The proposed method is designed to determine an appropriate number of clusters, in addition to the actual peaks, automatically. Experiments confirm that the proposed approach yields desirable results, demonstrating that it can find an appropriate set of clusters for a set of well-known benchmark images. (C) 2020 Elsevier Ltd. All rights reserved.
