Browsing by Author "Zertal, Soumia"

Now showing 1 - 2 of 2
  • No Thumbnail Available
    Item
    Cloud and Virtualization
    (University of Oum El Bouaghi, 2026) Zertal, Soumia
    This document is a course support for the subject entitled "Cloud and Virtualization ", taught in the Department of Mathematics and Computer Science at the University of Oum El Bouaghi and intended for second year Master's students in computer science, specializing in distributed architecture. The objective of this course is to provide students with: - Understanding of the principles of virtualization; - Manipulating the different concepts of virtualization through practical tools; - Discovering the concept of Cloud Computing and its applications; - Knowledge of the most well-known Cloud platforms and the ability to manipulate the services offered by at least one of these platforms. To achieve the stated objective, we made every effort to approach this work from multiple perspectives and synthesized the most relevant information using a variety of sources (books, lecture notes, articles, websites, etc.), while adhering to the official framework defined by the Ministry of Higher Education and Scientific Research. To make the course more engaging, we added practical sessions designed to facilitate the application of the theoretical concepts covered in the lectures. Nevertheless, we are aware that this document will remain partial, incomplete, and not exhaustive. For this reason, we strive to ensure continuous updates, with the aim of enriching its content. Therefore, we would be grateful if readers would point out any errors or offer suggestions in this regard
  • No Thumbnail Available
    Item
    Deep learning
    (University of Oum El Bouaghi, 2026) Zertal, Soumia
    This Deep Learning course handbook represents a comprehensive and rigorous exploration of one of artificial intelligence's most transformative disciplines, specifically designed for first-year Master's students specializing in Artificial Intelligence. The curriculum provides systematic coverage of both foundational principles and cutting-edge architectures that define contemporary deep learning practice. The course adopts a carefully structured pedagogical progression, beginning with theoretical foundations including neural network mathematics, activation functions, and backpropagation mechanisms, before advancing through specialized architectures: convolutional neural networks for visual data processing, recurrent networks (RNNs, LSTMs, GRUs) for sequential and temporal analysis, and generative models (GANs and VAEs) for data synthesis and augmentation. Each chapter integrates mathematical rigor with practical implementation, featuring real-world applications spanning computer vision, natural language processing, biomedical signal analysis, and autonomous systems. Students will develop essential competencies in designing, implementing, training, and critically evaluating deep neural networks while navigating architectural tradeoffs for domain-specific applications. The course emphasizes hands-on learning through exercises, projects, and implementation using modern frameworks, fostering both technical proficiency and conceptual intuition. Prerequisites include solid foundations in linear algebra, calculus, probability theory, and programming (preferably Python), along with basic machine learning concepts. Beyond technical mastery, the course instills critical awareness of ethical dimensions; fairness, transparency, privacy, societal impact, and environmental costs; that accompany powerful AI technologies. As future practitioners shaping artificial intelligence's trajectory, students are encouraged to approach these transformative capabilities with both excellence and responsibility, cultivating habits of continuous learning, rigorous thinking, and ethical innovation in a rapidly evolving field.