Altınbaş University Institutional Repository
Not a member yet
5805 research outputs found
Sort by
Enhancing cybersecurity and frequency control efficiency in AC islanded microgrids: a distributed approach with a stochastic models
No full textThis paper explores the vulnerability of AC islanded microgrids to cyber-attacks, focusing on communication links and local controls. Decentralized control methods are proposed to enhance security, with a detailed analysis of false data injection (FDI) and denial of service (DoS) attacks. Approaches to address DoS attacks include consensus of multiagent systems and risk-sensitive control. The study proposes a lightweight prevention model to mitigate FDI, featuring a cooperative stochastic control system with a communication-based controller. In contrast to existing distributed control methods, the proposed approach offers advantages in synchronizing frequency restoration, ensuring accurate power sharing through sparse communication networks, and reducing control update frequency. The study contributes to heightened system security, addresses limitations of centralized control methods, and integrates stationary proportional resonant primary control and intelligent secondary control based on the dragonfly algorithm for improved microgrid performance under various operating conditions
Develop a robust computer network architecture that is resistant to unauthorized access by using machine learning methodologies
No full textAs the use of IT spreads rapidly into new areas, the necessity to ensure the security of these
systems has grown. Cyberattacks have also become much more sophisticated as a result of
the widespread availability of information technology. Consequently, traditional security
measures like SIDS have failed to identify new types of assaults. Intrusion Detection
Systems (IDS) make it possible to track and gather harmful data inside a network. The
majority of IDSs rely on signatures to identify potential threats. They use a set of rules—
either manually entered by the administrator or created automatically by the system—to
identify and respond to known threats. To maintain the availability of services at all times,
network security experts focus on both preventing and responding to intrusion attempts. To
find and categorize suspicious actions, security professionals employ tools like IDS. Hence,
to protect privacy, security, and the ongoing delivery of services, it is crucial that the IDS
continually keeps up-to-date with the most recent intrusion attack signatures. Important
factors to consider while evaluating IDS performance are its speed and its capacity to learn
new assaults. This study demonstrates how several Machine Learning techniques may be
evaluated using the Knowledge Discovery and Data Mining (KDD) dataset, which is also
called Knowledge Discovery in Databases. The primary focus is on creating a
comprehensive and representative dataset for experimentation, with a strong emphasis on
KDD. For this analysis, we have chosen to use the K-Nearest Neighbor (KNN) and Multilayer Perceptron (MLP) classifiers. The KNN classifier has shown the highest accuracy
in recognizing and classifying all types of KDD dataset attacks (DOS, R2L, U2R,
NORMAL, and PROBE), both for binary class (NORMAL vs. ABNORMAL) and multiclass scenarios. The experimental findings utilizing the proposed KNN and MLP models
showed that the accuracy of binary classification using KNN and MLP was 99% and 97%
respectively. Furthermore, the multi-class classification produced improved results
compared to earlier work, with reported high-level accuracies of 92% and 87% respectively.
This thesis investigates the effectiveness of using deep learning techniques, namely MNN
and MLP designs, for network flow-based intrusion detection.BT kullanımı hızla yeni alanlara yayıldıkça, bu sistemlerin güvenliğinin sağlanması
gerekliliği de arttı. Bilgi teknolojisinin yaygınlaşmasının bir sonucu olarak siber saldırılar
da çok daha karmaşık hale geldi. Sonuç olarak, SIDS gibi geleneksel güvenlik önlemleri
yeni saldırı türlerini tespit etmekte başarısız oldu. İzinsiz Giriş Tespit Sistemleri (IDS), bir
ağ içindeki zararlı verileri izlemeyi ve toplamayı mümkün kılar. IDS'lerin çoğunluğu
potansiyel tehditleri tanımlamak için imzalara güveniyor. Bilinen tehditleri tanımlamak ve
bunlara yanıt vermek için yönetici tarafından manuel olarak girilen veya sistem tarafından
otomatik olarak oluşturulan bir dizi kural kullanırlar. Hizmetlerin her zaman
kullanılabilirliğini korumak için ağ güvenliği uzmanları, izinsiz giriş girişimlerini hem
önlemeye hem de bunlara yanıt vermeye odaklanır. Şüpheli eylemleri bulmak ve
sınıflandırmak için güvenlik uzmanları IDS gibi araçlar kullanır. Bu nedenle gizliliği,
güvenliği ve hizmetlerin devam eden sunumunu korumak için IDS'nin sürekli olarak en son
izinsiz giriş saldırısı imzalarıyla güncel kalması çok önemlidir. IDS performansını
değerlendirirken dikkate alınması gereken önemli faktörler, hızı ve yeni saldırıları öğrenme
kapasitesidir. Bu çalışma, Veritabanlarında Bilgi Keşfi olarak da adlandırılan Bilgi Keşfi ve
Veri Madenciliği (KDD) veri kümesi kullanılarak çeşitli Makine Öğrenimi tekniklerinin
nasıl değerlendirilebileceğini göstermektedir. Öncelikli odak noktası, KDD'ye güçlü bir
vurgu yaparak deneyler için kapsamlı ve temsili bir veri kümesi oluşturmaktır. Bu analiz için
K-En Yakın Komşu (KNN) ve Çok Katmanlı Algılayıcı (MLP) sınıflandırıcılarını
kullanmayı seçtik. KNN sınıflandırıcı, hem ikili sınıf (NORMAL vs. ABNORMAL) hem de çok sınıflı senaryolar için tüm KDD veri kümesi saldırı türlerini (DOS, R2L, U2R,
NORMAL ve PROBE) tanıma ve sınıflandırmada en yüksek doğruluğu göstermiştir.
Önerilen KNN ve MLP modellerini kullanan deneysel bulgular, KNN ve MLP kullanılarak
yapılan ikili sınıflandırmanın doğruluğunun sırasıyla %99 ve %97 olduğunu göstermiştir.
Ayrıca, çok sınıflı sınıflandırma, sırasıyla %92 ve %87'lik yüksek düzeyde doğruluk
oranlarıyla daha önceki çalışmalara kıyasla daha iyi sonuçlar üretti. Bu tez, ağ akışı tabanlı
izinsiz giriş tespiti için derin öğrenme tekniklerini, yani MNN ve MLP tasarımlarını
kullanmanın etkinliğini araştırmaktadır
Detection of Animals and humans in forest fires using Yolov8
No full text- The study uses the YOLOv8 deep learning algorithm to detect fire, smoke, humans, and animals in outdoor images. The importance of forests in protecting the biosphere is emphasized, and forest fires are identified as a major risk to the environment and living beings. The researchers created a custom dataset of outdoor images and manually annotated them. The YOLOv8 model was trained on this dataset, and its overall performance was evaluated, with varying results for different object classes. The study identified areas for improvement in the model's ability to detect small instances of fire and smoke and differentiate between animals and humans. The impact of image quality on the model's performance was also highlighted. Overall, the study provides a comprehensive evaluation of YOLOv8's performance in detecting outdoor objects and identifies areas for improvement
Farklı metotlarla yapılan arayüz mine aşındırmalarında uygulama başarısına çapraşıklığın etkisinin 3B tarayıcılar aracılığıyla In Vitro değerlendirilmesi
No full textAMAÇ: Çapraşıklığın mine arayüz aşındırma uygulama başarısına etkisinin üç boyutlu (3B)
tarayıcı aracılığıyla tespiti amaçlanmıştır.
GEREÇ VE YÖNTEM: Bu çalışmada iki farklı aşındırma miktarında (0,3 mm ve 0,5 mm)
aşındırılmak üzere 264 termoplastik mandibular keser diş (Frasaco GmbH, Tettnang,
Almanya), aşındırma materyali özelliklerine göre yirmi dört alt gruba ayrılmıştır. Tek ve çift
taraflı olmak üzere elmas disk, manuel abrazivler, motor destekli abrazivler aşındırma
uygulamalarında kullanılmıştır. Termoplastik dişler aşındırma öncesi ve sonrası 3B tarayıcı
aracılığıyla taranarak kaydedilmiştir. Kaydedilen 3B modeller Geomagic Kontrol X (3D
Systems, Amerika Birleşik Devletleri) programı üzerinde çakıştırılarak 3B karşılaştırma
analiz verileri aracılığıyla hedeflenen aşındırmaya ulaşma miktarı, aşınma miktarı ve tepe
noktası bölgesi, hata miktarı ve tepe noktası bölgesi, çapraşık dişlerde yapılan aşındırmada
bukkale taşma miktarı verileri her diş için raporlanmıştır.
BULGULAR: Materyale göre hedeflenene ulaşma miktarları istatistiksel olarak anlamlı
farklılık göstermemektedir (p>0,05). Yerleşime göre hedeflenene ulaşma miktarları çapraşık
dişlerde yüksek olarak istatistiksel anlamlı farklılık göstermektedir (p=0,001; p<0,01). Hata
miktarı ile aşındırma miktarı arasında istatistiksel olarak anlamlı bir ilişki bulunamamıştır
(p>0,05). Yöne göre hata miktarları çift yönlü apareylerde istatistiksel olarak anlamlı
derecede yüksek bulunmuştur (p=0,001; p<0,01). Materyale göre bukkale taşma miktarı manuel abraziv materyalinin bukkale taşma miktarının, elmas Disk ve motor destekli abraziv
materyallerinden yüksek olması anlamlı bulunmuştur. Diğer materyaller arasında
istatistiksel olarak anlamlı bir farklılık bulunmamıştır (p>0,01). Çapraşık dişlerde; yöne göre
bukkale taşma miktarı tek taraflı materyallerde daha yüksek bulunmuştur (p=0,001; p<0,01).
Çapraşık dişlerde motor destekli abraziv materyalinde; yöne göre bukkale taşma miktarı tek
taraflı olanlarda daha yüksek bulunmuştur (p=0,001; p<0,01). Elmas disk ve manuel
abrazivlerde yöne göre bukkale taşma miktarı yöne göre anlamlı farklılık göstermemiştir.
Yerleşime göre aşınma tepe bölgesi sıralı dişlerde %92,43 oranında, çapraşık dişlerde
%87,11 oranında orta bölgede konumlanmıştır. Hedeflenen miktara göre aşınma tepe bölgesi
0,3 mm hedeflenen dişlerde %86,37; 0,5 mm hedeflenenlerde %93,19 oranında orta
bölgededir. Sıralı dişlerde hata tepe noktası %100; çapraşık dişlerde %94,85 oranında orta
bölgededir. Hata tepe bölgesi çapraşık dişlerde %5,15 oranında lingualde konumlanmıştır.
SONUÇ: Çapraşıklık mine arayüz aşındırmasında hedeflenen miktara ulaşımı arttırmakta
ancak aşınma bölgesini çapraşık dişlerinin pozisyonuna göre idealden uzaklaştırmaktadır.
Çapraşık dişlerde yapılan aşındırmalarda çift taraflı aşındırma materyali kullanımı hata
miktarını arttırmaktadır. Aşındırma etkinliğinde materyaller arası bir üstünlük yoktur ancak
kullanılan materyalin cinsine ve tek taraflı olup olmadığına bağlı olarak çapraşık dişlerde
uygulanan aşındırmanın bukkale taşma miktarı değişmektedir.OBJECTIVE: It was aimed to determine the effect of crowding on the success of
interproximal enamel reduction (IPR) using a three-dimensional (3D) scanner.
MATERIALS AND METHODS: In this study, 264 thermoplastic mandibular incisors
(Frasaco GmbH, Tettnang, Germany) were divided into twenty-four subgroups according to
the abrasive material properties, to be abraded at two different abrasion amounts (0.3 mm
and 0.5 mm). Single and double-sided diamond discs, manual abrasives, and motor-assisted
abrasives have been used in abrasion applications. Thermoplastic teeth were scanned and
recorded using a 3D scanner before and after IPR. The recorded 3D models were
superimposed on the Geomagic Control X.
RESULTS: The amount of achieving the planned IPR does not show a statistically significant
difference depending on the material (p>0.05). The amount of reaching the planned IPR
according to the placement shows a statistically significant difference, which is higher in
crowded teeth (p=0.001; p<0.01). No statistically significant relationship was found between
the amount of error and the amount of IPR (p>0.05). Error amounts according to onesided/double-sided were found to be statistically significantly higher in double-sided
appliances (p=0.001; p<0.01). According to the material, it was found significant that the
amount of buccal overflow of manual abrasive material was higher than that of diamond disc
and motor-assisted abrasive materials. There was no statistically significant difference between other materials (p>0.01). In crowded teeth; depending on the sides, the amount of
buccal overflow was found to be higher in one-sided materials (p=0.001; p<0.01). Motorassisted abrasive material for crowded teeth; depending on the direction, the amount of
buccal overflow was found to be higher in one-sided materials (p=0.001; p<0.01). The
amount of buccal overflow did not differ significantly depending on the sides in diamond
disc and manual abrasives. According to the placement, the top reduction area is located in
the middle region with a rate of 92.43% in aligned teeth and 87.11% in crowded teeth.
According to the planned IPR amount, 86.37% of the teeth with a reduction peak area of 0.3
mm; For those planned IPR 0.5 mm, 93.19% of the time it is in the middle region. Error peak
for aligned teeth is 100%; In crowded teeth, 94.85% of the time it is in the middle region.
The error peak area is located lingually in 5.15% of crowded teeth.
CONCLUSION: Crowding increases the rate of achieving the targeted amount of IPR, but
it moves the reduction area away from the ideal depending on the position of the crowded
teeth. The use of double-sided abrasive material in IPR on crowded teeth increases the
amount of error. There is no superiority between materials in IPR effectiveness, but the
amount of IPR applied to crowded teeth varies depending on the type of material used and
whether it is one-sided or not
Integrating BERT for nuanced sentiment analysis: a detailed examination of diverse textual datasets
Get PDFThe rapid emergence and growth in the number of digital communication platforms have resulted in a previously unimaginable volume of text data suitable for sentiment analysis. Public sentiment extraction and interpretation definitive in various areas of marketing, politics, and customer service sectors. Existing approaches prove inadequate to address language complexity and subtlety. Consequently, more advanced analytical tools are required. Therefore, this study employs a BERT model in performing sentiment analysis on textual data associated with ChatGPT, an AI conversational tool OpenAI develops. The datasets utilized in this research are three, which are Data Collection, ChatGPT Sentiment Analysis Dataset, and ChatGPT App Reviews Dataset. They range from topics such as tweets to comprehensive reviews and thus provide diverse backgrounds for testing the model's applicability. The BERT model is chosen for its strong processing abilities, especially due to its capacity to understand text nuances. It applies sentiment analysis in six training epochs, and the results indicated a model competent in accurate sentiment classification, irrespective of language and sentiment uniqueness. The overall performance was ousttanding, as evidenced by the model's accuracy and its classification precision in both datasets. These findings have the potential to make the BERT model dominant in all sentiment analysis tasks. The analysis would prove instrumental in providing in-depth public opinion and sentiments made possible by the model's credible results. In this stutdy, it is clear that BERT models can significantly improve machine learning towards better understanding and interacting in human language. These results also provide a basis for further research into these modalities across real-time applications or further modification that would develop the model to be exposed to different kinds of data
Design and develop function for research based application of intelligent internet-of-vehicles model based on fog computing
Get PDFThe fast growth in Internet-of-Vehicles (IoV) applications is rendering energy efficiency management of vehicular networks a highly important challenge. Most of the existing models are failing to handle the demand for energy conservation in large-scale heterogeneous environments. Based on Large Energy-Aware Fog (LEAF) computing, this paper proposes a new model to overcome energy-inefficient vehicular networks by simulating large-scale network scenarios. The main inspiration for this work is the ever-growing demand for energy efficiency in IoVmost particularly with the volume of generated data and connected devices. The proposed LEAF model enables researchers to perform simulations of thousands of streaming applications over distributed and heterogeneous infrastructures. Among the possible reasons is that it provides a realistic simulation environment in which compute nodes can dynamically join and leave, while different kinds of networking protocols-wired and wireless-can also be employed. The novelty of this work is threefold: for the first time, the LEAF model integrates online decision-making algorithms for energy-aware task placement and routing strategies that leverage power usage traces with efficiency optimization in mind. Unlike existing fog computing simulators, data flows and power consumption are modeled as parameterizable mathematical equations in LEAF to ensure scalability and ease of analysis across a wide range of devices and applications. The results of evaluation show that LEAF can cover up to 98.75% of the distance, with devices ranging between 1 and 1000, showing significant energy-saving potential through A wide-area network (WAN) usage reduction. These findings indicate great promise for fog computing in the future-in particular, models like LEAF for planning energy-efficient IoV infrastructures
Design and simulation of a PV based smart grid for predictive energy generation
Get PDFThe successful implementation of PV-based smart grids necessitates a supportive policy and
regulatory framework that encourages renewable energy adoption, incentivizes grid
integration, and ensures fair compensation for stakeholders. This research aims to identify
the policy gaps and regulatory challenges hindering the deployment of PV-based smart grids
and propose recommendations for policy reforms that promote their widespread adoption.
By addressing the policy and regulatory implications, this research strives to create an
enabling environment that fosters the growth of PV-based smart grids and renewable energy
integration. the motivations driving this thesis encompass environmental sustainability,
energy security and resilience, technological advancements, efficient energy management,
and policy and regulatory implications. By addressing these motivations, this research aims
to contribute to the advancement of PV-based smart grids for predictive energy generation,
fostering a sustainable, reliable, and cost-effective energy management system for a greener
future
Multisource data framework for prehospital emergency triage in real-time IoMT-based telemedicine systems
No full textBackground and objective: The Internet of Medical Things (IoMT) has revolutionized telemedicine by enabling the remote monitoring and management of patient care. Nevertheless, the process of regeneration presents the difficulty of effectively prioritizing the information of emergency patients in light of the extensive amount of data generated by several integrated health care devices. The main goal of this study is to be improving the procedure of prioritizing emergency patients by implementing the Real-time Triage Optimization Framework (RTOF), an innovative method that utilizes diverse data from the Internet of Medical Things (IoMT).
Methods: The study's methodology utilized a variety of Internet of Medical Things (IoMT) data, such as sensor data and texts derived from electronic medical records. Tier 1 supplies sensor and textual data, and Tier 3 imports textual data from electronic medical records. We employed our methodologies to handle and examine data from a sample of 100,000 patients afflicted with hypertension and heart disease, employing artificial intelligence algorithms. We utilized five machine-learning algorithms to enhance the accuracy of triage.
Results: The RTOF approach has remarkable efficacy in a simulated telemedicine environment, with a triage accuracy rate of 98%. The Random Forest algorithm exhibited superior performance compared to the other approaches under scrutiny. The performance characteristics attained were an accuracy rate of 98%, a precision rate of 99%, a sensitivity rate of 98%, and a specificity rate of 100%. The findings show a significant improvement compared to the present triage methods.
Conclusions: The efficiency of RTOF surpasses that of existing triage frameworks, showcasing its significant ability to enhance the quality and efficacy of telemedicine solutions. This work showcases substantial enhancements compared to existing triage approaches, while also providing a scalable approach to tackle hospital congestion and optimize resource allocation in real-time. The results of our study emphasize the capacity of RTOF to mitigate hospital overcrowding, expedite medical intervention, and enable the creation of adaptable telemedicine networks. This study highlights potential avenues for further investigation into the integration of the Internet of Medical Things (IoMT) with machine learning to develop cutting-edge medical technologies
Investigation of the Relationship Between Temperature and Volume Fraction on the Mechanical Properties of a Polymeric Composite Material
No full text2nd International Conference on Engineering and Science to Achieve the Sustainable Development Goals, ICASDG 2023 -- 9 July 2023 through 10 July 2023 -- Hybrid, Tabriz -- 197984In this research, the properties of composite materials and the influence of volume fraction and temperature on these properties were studied, samples consisting of polyester as a matrix reinforced by glass fibers were prepared with different volume fractions (10%, 20%, 30%), After that, the mechanical properties of the prepared samples were tested under different temperatures (room temperature, 60 °C and 80 °C), the room temperature equal 23 °C. From the obtained results from the mechanical tests, the mechanical properties improve and increase with the increasing the volume fraction. It has been found that the hardness and compressive strength decrease when the temperature is increased, while the tensile strength, flexural resistance and impact resistance increase when the temperature is increased to 60 °C, but decrease when the temperature is increased to 80 °C, and this means that the tensile, flexural and impact resistance increase to a certain extent when increasing temperature, but it starts to drop when this limit is exceeded. © 2024 American Institute of Physics Inc.. All rights reserved
Türkiye’deki diş hekimliğinde uzmanlık eğitimi giriş sınavı sorularına ilişkin ChatGPT ve Bard’ın Bloom’un revize edilmiş taksonomisine dayalı performansı
Get PDFObjective: This study aimed to compare the performance of chat generative pretrained trans- former (ChatGPT) (GPT-3.5) and Bard, 2 large language models (LLMs), through multiple-choice dental specialty entrance examination (DUS) questions.
Methods: Dental specialty entrance examination questions related to prosthodontics and oral and dentomaxillofacial radiology up to 2021, excluding visually integrated questions, were prompted into LLMs. Then the LLMs were asked to choose the correct response and specify Bloom’s taxonomy level. After data collection, the LLMs’ ability to recognize Bloom’s taxonomy levels and the correct response rate in different subheadings, the agreement between LLMs on correct and incorrect answers, and the effect of Bloom’s taxonomy level on correct response rates were evaluated. Data were analyzed using McNemar, Chi-square, and Fisher–Freeman–Halton
exact tests, and Yate’s continuity correction and Kappa agreement level were calculated (P < .05).
Results: Notably, the only significant difference was observed between ChatGPT’s correct answer rates for oral and dentomaxillofacial radiology subheadings (P = .042; P < .05). For total prosth- odontic questions, ChatGPT and Bard achieved correct answer rates of 35.7% and 38.9%, respec- tively, while both LLMs achieved a 52.8% correct answer rate for oral and dentomaxillofacial
radiology. Moreover, there was a statistically significant agreement between ChatGPT and Bard on correct and incorrect answers. Bloom’s taxonomy level did not affect the correct response rates significantly.
Conclusion: The performance of ChatGPT and Bard did not demonstrate a reliable result on DUS questions, but considering rapid advancements in these LLMs, this performance gap will probably be closed soon, and these LLMs can be integrated into dental education as an interactive tool.Amaç: Bu çalışmanın amacı, iki büyük dil modeli (LLM) olan ChatGPT (GPT-3,5) ve Bard’ın Diş Hekimliğinde Uzmanlık Eğitimi Giriş Sınavındaki (DUS) çoktan seçmeli sorular üzerindeki perfor- mansını karşılaştırmaktır.
Yöntemler: Görsel içerikli sorular hariç olmak üzere, 2021 yılına kadar olan protetik diş tedavisi ve ağız, diş ve çene radyolojisi ile ilgili DUS soruları LLM’lere sorulmuştur. Daha sonra LLM’lerden doğru yanıtı seçmeleri ve Bloom’un taksonomi düzeyini belirtmeleri istenmiştir. Veriler toplandık- tan sonra, LLM’lerin Bloom taksonomi düzeylerini belirleyebilme becerileri ve farklı alt başlıklardaki doğru yanıt oranları, LLM’ler arasında doğru ve yanlış yanıtlara ilişkin uyumu ve Bloom taksonomi düzeyinin doğru yanıt oranları üzerindeki etkisi değerlendirilmiştir. Veriler Mc Nemar, Ki-kare ve Fisher Freeman Halton Exact testleri kullanılarak analiz edilmiştir, Yate’s Continuity Düzeltmesi ve Kappa uyum düzeyi hesaplanmıştır (P < .05).
Bulgular: ChatGPT’nin doğru cevap oranları arasında tek anlamlı fark ağız, diş ve çene radyolojisi alt başlıkları arasında gözlenmiştir (P: .042; P < .05). Toplam protez soruları için ChatGPT ve Bard sırasıyla %35,7 ve %38,9 oranında doğru cevap verirken, her iki LLM de ağız, diş ve çene radyolojisi için %52,8 oranında doğru cevap vermiştir. Ayrıca, ChatGPT ve Bard arasında doğru ve yanlış cevaplar konusunda istatistiksel olarak anlamlı bir uyum saptanmıştır. Bloom’un taksonomi düzeyi doğru yanıt oranlarını anlamlı derecede etkilememiştir.
Sonuç: ChatGPT ve Bard, DUS soruları üzerinde güvenilir bir performans göstermemiştir, ancak LLM’lerdeki hızlı gelişmeler göz önünde bulundurulduğunda, performans açıkları muhtemelen yakında kapanacak ve bu LLM’ler interaktif bir araç olarak diş hekim- liği eğitimine entegre edilebilecektir