Master Informatique
Permanent URI for this collectionhdl:123456789/1389
Browse
223 results
Search Results
Item Autism Detection Using Machine Learning(2024-09-23) Youcef Kouadria; Sakhraoui ZineddineThis research investigates the use of machine learning, specifically deep learning, for the detection of Autism Spectrum Disorder (ASD) using facial image data. By employing the VGG16 model with transfer learning, the project achieved high classification accuracy, demonstrating the potential of machine learning in supporting early ASD diagnosis. The use of automated image analysis provides a non-invasive, scalable solution that could complement traditional diagnostic methods. The results indicate that machine learning can significantly contribute to healthcare by enabling quicker and more accurate ASD diagnoses. Future improvements could include the use of larger datasets and multimodal data such as voice and behavioral analysisItem Classification automatique de la maladie de Parkinson à partir de la voix(Université de Bordj Bou Arreridj Faculty of Mathematics and Computer Science, 2024) Dahili, Ahlem; Benmessahel, SamiraAbstract Speech analysis is a promising approach for early and automated diagnosis of Parkinson's disease. This non-invasive and inexpensive method relies on the characteristic voice changes of the disease, present from the early stages, to identify patients. Automated systems based on artificial intelligence can analyze these voice changes and effectively discriminate Parkinson's disease patients from healthy subjects. Despite challenges such as voice variability and background noise, speech analysis has great potential to improve the diagnosis and management of Parkinson's disease. Ongoing research aims to refine this technology and make it a valuable tool for improving the quality of life for patients.Item Suivre les rumeurs dans les réseaux sociaux(Université de Bordj Bou Arreridj Faculty of Mathematics and Computer Science, 2024) LAALAOUI, Moun; LAHRI, SarahThe emergence of the Internet has transformed global communications, making information accessible at unprecedented speeds. With the emergence of social media, people can now connect, share and interact immediately with a variety of content. However, the ease with which information can be exchanged also facilitates the rapid spread of unverified rumors and false information. In this work, we aim to track and detect rumors, and to this end, we will present a model based on a deep learning approach using LSTM and RNN algorithms in order to obtain the best possible classification and more accurate and valid results.Item L'autonomie de l'apprentissage : le e-learning à travers la plateforme Moodle. Cas des étudiants de première année licence français Université Mohamed El Bachir El Ibrahimi BBA(Université de Bordj Bou Arreridj Faculty of Mathematics and Computer Science, 2024) -Dounia, MEZHOUD; - Amel BEN, REDOUANE; - Maria, RAHMANIItem L’Apprentissage Automatique Pour La Prédiction De Lien Dans Les Réseaux Complexes(Université de Bordj Bou Arreridj Faculty of Mathematics and Computer Science, 2024) BOUABDALLAH, Maroua; DRIAI, IbtissemThis work explores graph theory concepts to model and analyze complex networks with an emphasis on the use of machine learning. Methods examined include similarity measures based on common neighbors, measures based on the length of paths, We also evaluated the effectiveness of different classification algorithms, such as Support Vector Machine (SVM), K Nearest Neighbors (KNN)…Our results show that certain combinations of these methods and algorithms make it possible to obtain accurate predictions of link classes in complex networks, thus opening new perspectives for their analysis and application in various fielItem Approche Multifacette pour la Maladie du Foie : Prédiction, Méta-Classification et Simulation de la Migration entre Stades(Université de Bordj Bou Arreridj Faculty of Mathematics and Computer Science, 2024) Nouioua, Imene; Nouioua Ratiba, Ratibahronic diseases, especially those affecting the liver, pose a major challenge to global heal thcare systems. In this dissertation, we have explored various aspects from prediction to simula ting the migration between stages of these chronic conditions. Utilizing advanced data analysis and machine learning techniques, our study focuses on four key aspects : improving predic tion, feature selection, model optimization, and meta-classification, along with simulating the migration between disease stages for preventive purposes. At each stage, rigorous experiments were conducted to validate our methodology. The results confirm the crucial importance of prediction in anticipating disease progression, as well as the effectiveness of feature selection and model optimization in enhancing prediction performance. Meta-classification, by combi ning predictions from different models, enhances result reliability. Furthermore, simulating the migration between stages provides a better understanding of disease progression dynamicsItem Développement et Évaluation d'un Système de Reconnaissance Faciale Basé sur le Classificateur Deep Rule-Based (DRB)(Université de Bordj Bou Arreridj Faculty of Mathematics and Computer Science, 2024) Belkhiri, khawla; Sahli maissa, maissaFacial recognition systems have become one of the most widely used systems in the fields of security and surveillance. Despite significant technological advancements, their performance is still affected by changes in shooting conditions, such as lighting and modifications to facial features caused by aging or different facial expressions. This thesis aims to improve the performance of facial recognition systems using a new DRB classifier. The proposed solutions have led to significant improvements compared to other classifiers (NN classifier, SVM). Image matching descriptors Gabor, LPQ, MBC, and IWBC were used in experiments applied to the ORL, 15 Yale, Face94, Face95, Face96, and Jaffe databases, which are among the most used in academic studies to compare the results and determine the degree of improvement achievedItem Prédiction des tumeurs cérébrales dans les images IRM par l’apprentissage profond(Université de Bordj Bou Arreridj Faculty of Mathematics and Computer Science, 2024) MERDJI, Saida; REBIAI, SoriaTThe accurate diagnosis of contemporary diseases heavily relies on the processing of medical images. This study introduces an interesting approach for automated detection of brain tumors from magnetic resonance imaging (MRI) using the deep learning model ResNet50. This model, renowned for its ability to extract complex features from images, is deployed to analyze brain MRI images and accurately identify the presence of tumors. The data used in this study include MRI images containing tumors. We compared our approach to other methods using criteria such as precision, recall, and F1 score. The proposed model, ResNet50, achieved a detection accuracy of 98%, demonstrating its effectiveness in detecting brain tumors from MRI images. These results highlight the potential of the ResNet50 model to improve early and accurate detection of brain tumors in images.Item Découverte de règles de fouille de motifs dans les bases de données transactionnelles(Université de Bordj Bou Arreridj Faculty of Mathematics and Computer Science, 2024) Miloudi, Fatima; Zohra Hammadi Moncef, Hammadi MoncefA major problem in data mining is High Utility Pattern Mining (HUPM), which seeks to find combinations of items that have a significant impact on a specific metric, such as sales, profits or customer satisfaction. Due to the growth in the volume of data in the field of Big Data, it is essential to design efficient algorithms to quickly extract these sets of high-value elements. In our study, we address the topic of finding high utility patterns in real transactional data bases. The objective is to discover very useful patterns in these bases. The utility of an item in the database represents its importance in relation to other items ; it can often be associated with the price of the item, but can also be defined by other criteria. Two algorithms were tested and applied on two real bases : the first from a pharmacy and the second containing purchases made in a fruit shop. This allows to extract two different forms of high-utility patterns : High Utility Itemsets (HUIs) and High Utility Association Rules (HARs)Item L’apprentissage profond pour la reconnaissance des macro-expressions(Université de Bordj Bou Arreridj Faculty of Mathematics and Computer Science, 2024) Merrouche, Said; Ben Merrouche, Imad el hakRecognition of human emotions, particularly through facial expressions, has recently garnered a lot of research attention. Advanced deep learning and machine learning techniques have been employed to analyze the CK+ database in order to better understand and identify emotions. In our experiments, we explored two primary methods for emotion detection. The first method involved machine learning techniques using algorithms such as k-nearest neighbors (K-NN) and support vector machines (SVM). The second method relied on deep learning using convolutional neural networks (CNN) and (DenseNet). This comparison allowed us to evaluate the effectiveness of traditional approaches versus modern techniques in the field of emotion recognition, providing us with deep insights into the relative performance of each