12 sonuçlar
Arama Sonuçları
Listeleniyor 1 - 10 / 12
Yayın Efficient estimation of Sigmoid and Tanh activation functions for homomorphically encrypted data using Artificial Neural Networks(Institute of Electrical and Electronics Engineers Inc., 2024) Harb, Mhd Raja Abou; Çeliktaş, BarışThis paper presents a novel approach to estimating Sigmoid and Tanh activation functions using Artificial Neural Networks (ANN) optimized for homomorphic encryption. The proposed method is compared against second-degree polynomial and Piecewise Linear approximations, demonstrating a minor loss in accuracy while maintaining computational efficiency. Our results suggest that the ANN-based estimator is a viable alternative for secure machine learning models requiring privacypreserving computation.Yayın A context-aware, AI-driven load balancing framework for incident escalation in SOCs(Institute of Electrical and Electronics Engineers Inc., 2025-08-12) Abuaziz, Ahmed; Çeliktaş, BarışSOCs face growing challenges in incident management due to increasing alert volumes and the complexity of cyberattacks. Traditional rule-based escalation models often fail to account for the workload of the analyst, the severity of the incident, and the organizational context. This paper proposes a context-aware, AI-driven load balancing framework for intelligent analyst assignment and incident escalation. Our framework leverages large language models (LLMs) with retrievalaugmented generation (RAG) to evaluate incident relevance and historical assignments. A reinforcement learning (RL)-based scheduler continuously optimizes incident-to-analyst assignments based on operational outcomes, enabling the system to adapt to evolving threat landscapes and organizational structures. Planned simulations in realistic SOC environments will compare the model with traditional rule-based models using metrics such as Mean Time to Resolution (MTTR), workload distribution, and escalation accuracy. This work highlights the potential of AIdriven approaches to improve SOC performance and enhance incident response effectiveness.Yayın Assessing ChatGPT's accuracy in dyslexia inquiry(Institute of Electrical and Electronics Engineers Inc., 2024) Eroğlu, Günet; Harb, Mhd Raja AbouDyslexia poses challenges in accessing reliable information, crucial for affected individuals and their families. Leveraging chatbot technology offers promise in this regard. This study evaluates the OpenAI Assistant's precision in addressing dyslexia-related inquiries. Three hundred questions commonly posed by parents were categorized and presented to the Assistant. Expert evaluation of responses, graded on accuracy and completeness, yielded consistently high scores (median=5). Descriptive questions scored higher (average=4.9568) than yes/no questions (average=4.8957), indicating potential response challenges. Statistical analysis highlighted the significance of question specificity in response quality. Despite occasional difficulties, the Assistant demonstrated adaptability and reliability in providing accurate dyslexia-related information.Yayın A metric-driven IT risk scoring framework: incorporating contextual and organizational factors(Institute of Electrical and Electronics Engineers Inc., 2025-09-24) Ünal, Nezih Mahmut; Çeliktaş, BarışRisk analysis is a critical process for organizations seeking to manage their cybersecurity posture effectively. However, traditional risk analysis frameworks, such as the Common Vulnerability Scoring System (CVSS), primarily evaluate technical impacts without incorporating organizational context and dynamic risk factors. This paper presents a metric-based risk analysis framework designed to provide a more adaptable and context-aware risk-scoring framework. The proposed model enables risk owners to define customized threat scenarios and dynamically adjust metric weights based on organizational needs. Unlike traditional approaches, our method integrates contextual parameters to improve the accuracy and relevance of risk calculations. Experimental evaluations demonstrate that the proposed framework enhances risk prioritization and provides more actionable insights for decision-makers. This study contributes to the field by addressing the limitations of existing risk analysis models and offering a systematic approach for cybersecurity risk management.Yayın Mahremiyeti koruyan, merkezi, hibrit film öneri sistemi: araçlar arası internet için bir yaklaşım(Institute of Electrical and Electronics Engineers Inc., 2025-08-15) Şimşek, Musa; Tüysüz Erman, AyşegülBu çalışmada, kullanıcı verilerinin gizliliğini korurken öneri doğrulu günü artırmayı hedefleyen, diferansiyel mahremiyet destekli hibrit bir öneri modeli sunulmuştur. Model mimarisi, Matris Çarpanlaması (MF), Çok Katmanlı Algılayıcı (MLP) ve Uzun Kısa Süreli Bellek (LSTM) ağlarını birleştirmektedir. Laplace mekanizmasına dayalı gürültü enjeksiyonu ile eğitim sürecinde diferansiyel mahremiyet sağlanmış ve ayrıca hiperparametre optimizasyonu uygulanmıştır. Model, kullanıcı film etkileşimlerini içeren MovieLens 100K veri kümesi üzerinde değerlendirilmiştir. Performans değerlendirmesi MSE, MAE ve NDCG metrikleriyle yapılmış; hiperparametre optimizasyonu ile MSE bazında yaklaşık %4 iyileşme sağlandığı, yüksek gizlilik düzeyinde ise doğrulukta yaklaşık %39 oranında bozulma yaşandığı gözlemlenmiştir.Yayın Çok ölçekli görsel benzerlik analizi ile oltalama saldırısı tespiti(Institute of Electrical and Electronics Engineers Inc., 2025-08-15) Kılıç, Bartu; Çeliktaş, BarışOltalama saldırıları teknolojinin gelişmesiyle günümüzün en yaygın siber güvenlik tehditlerinden biri haline gelmiştir. Bu çalışma, web sitelerinin ekran görüntülerini gelişmiş bir görsel benzerlik analizi yöntemiyle inceleyerek oltalama saldırılarını yüksek doğrulukla tespit eden bir yaklaşım sunmaktadır. Oltalama tespiti için önerilen yöntemde, algısal özütleme tabanlı çoklu çözünürlük analizi, akıllı ilgi bölgesi (ROI) tespiti ve çoklu metrik füzyonu gibi teknikler birleştirilerek yüksek doğrulukta tespit yapılabilmektedir. Veri seti, popüler bankacılık, e-posta ve sosyal medya platformlarının gerçek ve oltalama sayfalarından oluşan 23 gerçek ve 3 oltalama sayfası ekran görüntülerinden derlenmiştir. Yapılan testler, yöntemin %85 doğruluk oranı ile tekil metrik tabanlı yaklaşımlardan daha iyi performans gösterdiğini ortaya koymuştur. Dil bağımsız çalışan bu yöntem, URL ve HTML manipülasyonlarına karşı dayanıklıdır ve gerçek zamanlı oltalama tespiti için güçlü bir çözüm sunmaktadır.Yayın Comparing pre-trained and fine-tuned transformer-based models for sentiment analysis in Turkish comments in student surveys(Institute of Electrical and Electronics Engineers Inc., 2025-08-15) Pourjalil, Kajal; Ekin, Emine; Recal, FüsunStudent surveys are essential for evaluating teaching quality and course content, but analyzing open-ended responses is challenging due to their unstructured and multilingual nature. This study applies sentiment analysis to Turkish educational survey responses using three transformer-based models: SAVASY, DBMDZ BERT Base Turkish Cased, and XLM-RoBERTa Base. A labeled dataset of real-world student comments was used, with sentiment labels assigned using the Gemini AI tool to facilitate model fine-tuning. Evaluation metrics included accuracy, F1-score, precision, recall, and confidence scores. Results show that fine-tuning improves sentiment classification, effectively identifying positive, negative, and neutral sentiments. This highlights the value of transformer models in analyzing Turkish student feedback.Yayın Privacy-preserving cyber threat intelligence: a framework combining private information retrieval, federated learning, and differential privacy(Institute of Electrical and Electronics Engineers Inc., 2025-09-21) Çamalan, Emre; Çeliktaş, BarışThreat Intelligence Platforms (TIPs) are essential for sharing indicators of compromise (IoCs), but querying them can leak sensitive organizational data. We propose a privacy-preserving framework that combines Private Information Retrieval (PIR), Federated Learning (FL), and Differential Privacy (DP) to mitigate this risk. Our approach addresses both content-level and metadata-level privacy concerns while supporting collaborative learning across organizations. It ensures that sensitive query patterns remain hidden, local threat data never leaves organizational boundaries, and model updates are protected against inference attacks. The framework integrates with existing TIPs such as MISP and OpenCTI, requiring minimal operational changes. We implement a prototype using a simulated Abuse IP dataset and evaluate it on latency, accuracy, and communication overhead. The system supports private queries in under 300 ms and maintains over 95% model accuracy under DP noise. These results indicate that strong privacy can be achieved with minimal performance trade-offs, making the approach viable for real-world CTI environments.Yayın Retinal disease classification from bimodal OCT and OCTA using a CNN-ViT hybrid architecture(Institute of Electrical and Electronics Engineers Inc., 2025-09-21) Aydın, Ömer Faruk; Tek, Faik Boray; Turkan, YaseminRetinal diseases are the leading cause of vision impairment and blindness worldwide. Early and accurate diagnosis is critical for effective treatment, and recent advances in imaging technologies such as Optical Coherence Tomography (OCT) and OCT Angiography (OCTA), have enabled detailed visualization of the retinal structure and vasculature. By leveraging these modalities, this study proposes an advanced deep learning architecture called MultiModalNet for automated multi-class retinal disease classification. MultiModalNet employs a dual-branch design, where OCTA projection maps are processed through a ResNet101 encoder, and cross-sectional slices from the OCT volume (B-scans) are analyzed using a Vision Transformer (ViT-Large). The extracted features from both branches were fused and passed through the fully connected layers for the final classification. Evaluated on the 3-class OCTA-500 dataset, which includes Age-related Macular Degeneration (AMD), Diabetic Retinopathy (DR), and Normal cases, the proposed model achieved state-of-the-art classification accuracy of 94.59 percent, significantly o utperforming single-modality baselines. This result highlights the effectiveness of integrating vascular and structural information to improve the diagnostic performance. The findings suggest that hybrid multi-modal deep learning approaches can play a transformative role in computer-aided ophthalmology, enhancing both clinical decision-making and screening workflows.Yayın Self-supervised learning of 3D structure from 2D OCT slices for retinal disease diagnosis on UK biobank scans(Institute of Electrical and Electronics Engineers Inc., 2025-09-21) Nazlı, Muhammet Serdar; Turkan, Yasemin; Tek, Faik BorayThis study presents a self-supervised learning framework for retinal disease classification using Optical Coherence Tomography (OCT) scans. To balance the contextual richness of 3D volumes with the computational efficiency of 2D architectures, we introduce a quasi-3D input generation strategy. Each input is constructed by stacking three OCT slices, sampled from channel-specific Gaussian distributions centered on the volume midplane, and arranged in a standard three-channel 2D format compatible with existing pre-trained models. These quasi-3D images are used to pre-train a Vision Transformer (ViT-Base) via a Masked Autoencoder (MAE) with a shared masking pattern, encouraging the model to reconstruct masked regions by encoding anatomical continuity across slices. Pre-training is conducted on 10,000 unlabeled OCT volumes from the UK Biobank. The encoder is then fine-tuned on the OCTA-500 dataset for three-class and four-class retinal disease classification tasks, including macular degeneration and diabetic retinopathy. The model achieves 92.57% accuracy on the three-class task, matching the performance of RETFound while using over 150 times less pre-training data and a smaller backbone.
