DSpace at UMBB
The institutional repository of the University M'Hamed Bougara Of Boumerdes is a digital archive that contains the scientific output of the University. Dspace manages, preserves and provides access to the academic works of UMBB.
Communities in DSpace
Select a community to browse its collections.
Recent Submissions
تحليل النزاعات الدولية
(جامعة أمحمد بوقرة بومرداس : كلية الحقوق و العلوم السياسية -بودواو, 2025) حمياز, سمير
Comparative Evaluation of StyleGAN3-Based Augmentation Strategies for Enhanced Medical Image Classification
(CEUR-WS, 2025) Touazi, Faycal; Gaceb, Djamel; Tadrist, Amira; Bakiri, Sara
Deep learning models for medical image classification face significant challenges due to class imbalance and
the limited availability of annotated datasets, particularly for rare diseases. Traditional data augmentation
techniques, such as rotation, translation, etc., often fail to provide sufficient diversity to perform a good
classification for minor classes. To address this issue, various strategies have been explored, including
oversampling, undersampling, cost-sensitive learning, and synthetic data generation using generative
adversarial networks (GANs). In this study, we evaluate the impact of using a generative AI based
approaches and demonstrate that the most effective strategy is to combine synthetic augmentation with
traditional methods. Specifically, we employ StyleGAN3 to generate high-fidelity synthetic images that,
when integrated with traditional data-augmentation techniques, may improve the performance of deep
learning models on medical image classification. We validate our method on datasets, including COVID-19
chest X-rays and HAM10000. Experimental results show that this hybrid approach leads to an improvement
in classification accuracy, particularly for minority classes, surpassing standalone augmentation strategies.
Our findings highlight the potential of AI-driven synthetic data generation as a complementary solution to
traditional augmentation, offering a more balanced and diverse dataset for medical image analysis.
Exploring Multi-Channel GPS Receivers for Detecting Spoofing Attacks on UAVs Using Machine Learning
(Multidisciplinary Digital Publishing Institute, 2025) Mouzai, Mustapha; Riahla, Mohamed Amine; Keziou, Amor; Fouchal, Hacène
All current transportation systems (vehicles, trucks, planes, etc.) rely on the Global Positioning System (GPS) as their main navigation technology. GPS receivers collect signals from multiple satellites and are able to provide more or less accurate positioning. For civilian applications, GPS signals are sent without any encryption system. For this reason, they are vulnerable to various attacks, and the most prevalent one is known as GPS spoofing. The main consequence is the loss of position monitoring, which may increase damage risks in terms of crashes or hijacking. In this study, we focus on UAV (unmanned aerial vehicle) positioning attacks. We first review numerous techniques for detecting and mitigating GPS spoofing attacks, finding that various types of attacks may occur. In the literature, many studies have focused on only one type of attack. We believe that targeting the study of many attacks is crucial for developing efficient mitigation mechanisms. Thus, we have explored a well-known datasetcontaining authentic UAV signals along with spoofed signals (with three types of attacked signals). As a main contribution, we propose a more interpretable approach to exploit the dataset by extracting individual mission sequences, handling non-stationary features, and converting the GPS raw data into a simplified structured format. Then, we design tree-based machine learning algorithms, namely decision tree (DT), random forest (RF), and extreme gradient boosting (XGBoost), for the purpose of classifying signal types and to recognize spoofing attacks. Our main findings are as follows: (a) random forest has significant capability in detecting and classifying GPS spoofing attacks, outperforming the other models. (b) We have been able to detect most types of attacks and distinguish them
Enhancing Data Privacy in Intrusion Detection: A Federated Learning Framework With Differential Privacy
(John Wiley and Sons Ltd, 2025) Saidi, Ahmed; Khouri, A. Ouadoud
The rise of cyber threats has underscored the critical need for robust intrusion detection systems (IDS). While traditional approaches, including statistical, knowledge-based, and AI-driven methods, have been pivotal, they often face limitations such as data privacy concerns, scalability challenges, and low detection accuracy on unfamiliar threats. This paper addresses these issues by adopting a federated learning (FL) paradigm for collaborative intrusion detection, allowing data to remain local and enhancing privacy protection. The proposed solution integrates advanced encryption techniques and differential privacy to safeguard confidentiality while ensuring system scalability and adaptability. By introducing a robust separation of agents' roles and leveraging FL's decentralized architecture, the system overcomes the limitations of centralized learning, including single points of failure and communication overhead. Experimental results validate the proposed architecture, demonstrating significant improvements in performance and offering a promising direction for modern network security. This work not only highlights the potential of FL-based IDS but also explores the integration of distributed ledger technologies to further enhance trust and security. These findings contribute to the growing field of privacy-preserving computing and lay the groundwork for future innovations in scalable, secure, and efficient intrusion detection systems
Predicting Methane Hydrate Formation Temperature in the Presence of Diverse Brines Using Explainable Artificial Intelligence
(American Chemical Society, 2025) Nait Amar, Menad; Zeraibi, Noureddine; Alqahtani, Fahd Mohamad; Djema, Hakim; Benamara, Chahrazed; Saifi, Redha; Gareche, Mourad; Ghasemi, Mohammad; Merzoug, Ahmed
Thisstudy presents three advanced techniques, includingthe leastsquares support vector machine (LSSVM), categorical boosting (CatBoost),and cascaded forward neural network (CFNN), to model methane hydrateformation temperature (MHFT) across various brines under a wide pressurerange. Utilizing a comprehensive data set of nearly 1000 samples,the models underwent rigorous training and testing phases. Graphicalanalyses and statistical assessment confirmed the high accuracy ofthe implemented models, with the CFNN scheme outperforming the others,achieving a total root-mean-square error (RMSE) of 0.3569 and an R2 of 0.9977. Comparison with existing modelsfurther highlighted the CFNN model’s superior performance.Additionally, the Shapley Additive exPlanning (SHAP) method was employedto enhance the aspects related to predictions’ explainabilityby assessing the impact of different inputs on the outcomes. Lastly,the proposed model holds significant potential for advancing industrialand academic applications related to hydrate phenomena