Faculté des mathématiques et de l'informatique
Permanent URI for this communityhdl:123456789/17
Browse
3 results
Search Results
Item Modélisation de la diffusion de l’innovation dans les réseaux sociaux(University of Mohamed El Bachir El Ibrahimi - Bordj Bou Arréridj, 2026) Rima BenfredjRecently, the diffusion of innovation has seen a paradigm shift and emerged as a renewed and interesting field due to advances in artificial intelligence, behavioral modeling, and empirical simulation. Understanding adoption behavior is increasingly complex, as individuals’ decisions are influenced not only by innovation attributes and social pressure but also by psychological characteristics, especially personality traits. This thesis offers a personality-driven diffusion model that integrates Big Five personality framework (OCEAN) into the Rogers' Diffusion of Innovation theory. Using agent-based modeling (ABM), individuals are presented as agents having distinct profiles, which shapes their perception of innovation features (relative_advantage, compatibility, complexity, trialability, observability). The framework consists of four major phases: perception, communication, persuasion, and decision. The special feature of the present research is the two-phase modeling strategy: (1) an exploratory simulation with randomly generated values in order to understand the fundamental diffusion dynamics. (2) the model is applied empirically, by applying a hybrid BERT-Random Forest model to predict Twitter users' personality traits based on ChatGPT-related data, these extracted qualities are subsequently used to regenerate the values of the remaining model' attributes and simulate adoption processes. The findings demonstrate that actually adoption behavior is profoundly influenced by personality driven dimensions, resulting in more realistic adoption curves than traditional models. This approach opens up new perspectives for the behavioral study of innovation diffusionItem Deep Learning-based Anomaly Detection in Network Traffic Patterns(university of bordj bou arreridj, 2024) HEDJAM Lidia; BELOUAHRI AyaThe anomaly in network traffic is a crucial issue that can cause significant losses in network security and performance. This prompted us to undertake this work to detect these anomalies accurately and promptly using deep learning techniques. This thesis investigates the use of long short-term memory (LSTM) neural networks, one of the deep learning methods, to detect anomalies in network data flows. LSTMs are well suited to this task thanks to their ability to capture long-term temporal dependencies. Our approach is distinguished by its ability to detect complex and varied anomalies, thus improving the security and efficiency of computer networks. The results show a significant improvement over traditional methodsItem Using Multi-objective Meta-heuristics for Data Mining(university of bordj bou arreridj, 2025-01-05) Soumaia KAHLOULThe ability to extract knowledge from large datasets is essential for innovation and informed decision-making, a process known as knowledge extraction or data mining. Traditional methods often fall short in fully utilizing data potential, necessitating the development of new algorithms for better insights. This thesis explores an innovative approach by integrating deep learning with advanced feature selection techniques to improve the classification accuracy of COVID-19 cases from chest X-ray images. The dataset includes X-ray images categorized as COVID-19, pneumonia, and normal. We employ the Binary Multi-Objective Henry Gas Solubility Optimization Algorithm (B-MOHGSO) for feature selection and leverage models like AlexNet, VGG19, GoogleNet, and ResNet for feature extraction. Eight versions of B-MOHGSO were tested, with k-nearest neighbors (k-NN) as the classifier. The study highlights the significant impact of S-shaped and V-shaped transfer functions on binary transformations and classifier performance in high-dimensional medical imaging. Notably, B-MOHGSO algorithms, particularly those using V-shaped transfer functions, excelled in selecting relevant features while maintaining high accuracy. When combined with the VGG19 model and SVM classifier, B-MOHGSO significantly reduced the feature set without sacrificing performance. The application of B-MOHGSO in COVID-19 classification is crucial for identifying key features that enhance diagnostic processes and treatment strategies. By adapting MOHGSO for discrete optimization, this research aims to address the complexities of high-dimensional medical data and improve healthcare analytics outcomes.