Abdellah Bakhouyi

Background: WDBC is a small historical benchmark, and near-ceiling performance alone provides limited evidence of transportability. Methods: We evaluated five model families for discrimination, calibration, and paired statistical testing. We utilized sensitivity-oriented out-of-fold (OOF) thresholds and decision curve analysis (DCA) and conducted an independent TOMPEI-CMMD methodological replication (larger two-center mammography cohort with biopsy-confirmed diagnoses; final cohort: 1358 patients, 1380 breasts; held-out: 272 patients, 279 breasts). Results: Across 20 additional stratified WDBC partitions, ROC-AUC remained consistently high (model means 0.988–0.995), but the criterion-specific nominal leader changed across partitions. Held-out TOMPEI ROC-AUC ranged from 0.775 to 0.804. No model demonstrated statistically significant superiority in ROC-AUC or frozen-threshold balanced accuracy after patient-cluster paired bootstrap comparison and Holm correction. The independent methodological replication yielded more moderate discrimination than the WDBC benchmark. Frozen sensitivity-oriented OOF thresholds transported reasonably but not uniformly. Conclusions: Multi-dimensional evaluation is more defensible than nominal AUC ranking. Neither the WDBC benchmark results nor the TOMPEI methodological replication establish population screening performance or clinical deployment readiness; further prospective clinically representative validation is required.

Predictive analytics have become increasingly capable of delivering actionable and accessible feedback to enhance teacher performance to enhance student outcomes in higher education. This study introduces a supervised machine learning predictive model designed to forecast the duration required to complete a course in a video learning environment using a dataset of 8,665 statements from 490 students from National Higher School of Art and Design at Hassan II University in Casablanca over six academic years (2019-24). This paper analyzes decision trees (DT), random forest (RF), support vector machines (SVM), gradient boosting (GB), and linear regression (LR) techniques. The CMI-5 standard and JSON format are used to automatically transfer learning activity data from the learning management system (LMS) to the learning record store (LRS). The results indicate that DT, RF, and GB achieved 100 percent predictor accuracy.

Smart learning, a field marked by rapid evolution and innovation, leverages emerging technologies to address modern educational challenges, transforming teaching methodologies and driving significant progress. This study analyzes the impact of key technologies, including artificial intelligence (AI), the Internet of Things (IoT), big data, and generative AI, between 2010 and 2024 using bibliometric and content analysis methods. Drawing from Scopus and Web of Science (WoS) databases, it highlights how these innovations foster personalized learning environments and dynamic educational content. The findings reveal exponential growth in research on AI and the IoT in education since 2015, with major contributions from Chinese and American researchers. The study profiles influential researchers, leading institutions, and pioneering countries, offering insights into the evolving landscape of smart education. By examining trends and the interplay between technology and educational reform, the paper underscores the importance of data-driven strategies for designing and implementing adaptive learning systems. It also anticipates future challenges and opportunities, proposing a framework to guide ethical integration of these technologies into education, ensuring they enhance global learning outcomes in an increasingly digital world.

Smart education, a new field of technology related to education, has emerged as a unique response to current educational challenges. This is becoming increasingly important for academic progress and aligns with the transformative impact of technology. This study addresses the transformative impact of smart technologies on education, focusing on the integration of the internet of things, big data, and artificial intelligence. Through a bibliometric and content analysis based on Scopus and Web of Science databases, we identify the most active researchers, leading universities, and the countries that contribute most significantly to the field of smart education. The findings reveal a significant increase in related publications, highlighting the growing importance of these technologies in enhancing teaching and learning experiences. The study shows the advantages and challenges of adopting such technologies, providing insights into their practical applications and the future direction of educational innovations. Integrating smart technologies in education is crucial for improving quality of life and academic outcomes, necessitating further research and development to fully realize their potential. This research contributes to the understanding of technological impacts on education and supports the development of strategies for their effective implementation.

The relentless evolution of communication systems, driven by the demands of 5G and the impending 6G networks, necessitates heightened data rates and spectral efficiency. orthogonal frequency division multiplexing (OFDM), a form of multicarrier modulation employed in multi-input multi-output (MIMO) systems, stands as a pivotal technology. Yet, OFDM grapples with challenges, notably the peak-to-average power ratio (PAPR) issue. Selective mapping (SLM) has been a favored technique for mitigating PAPR in OFDM, albeit challenged by computational complexities in its pursuit of discovering optimal phase factors. This paper pioneers a transformative approach by integrating metaheuristic algorithms genetic algorithm (GA), particle swarm optimization (PSO), and the innovative fireworks algorithm (FWA) into SLM for PAPR reduction while minimizing computational complexity. Simulation results not only affirm the efficacy of SLM-based techniques but also spotlight the potential of metaheuristic algorithms in addressing PAPR challenges in modern communication systems. The study transcends single-antenna systems, extending to MIMO-OFDM systems based on WiMAX standards, validating the efficacy of these techniques in multi-antenna configurations. Crucially, the FWA, proposed for the first time in this paper, emerges as a robust candidate, striking an enviable balance between computational efficiency and performance, achieving a notable PAPR reduction with a favorable search number.

The current educational system radically changed during the last years and con-tinues to grow. The Smart education is now a typical feature of education issue from a new information and communication technologies and of the introduction of these new technologies in the institutional learning. The introduction of these technologies in education has been associated with a rise evolution of people's quality of life by improving teaching and learning. MOOCs, are some of the re-cent technologies, which have been introduced recently in the higher education sector. This article focuses on the level of intelligence introduced into the online teaching and learning process. Different MOOCs will be presented, compared and analyzed according to a comparative study of multicriteria analysis by apply-ing decision analysis methods; Analytical Hierarchy Process (AHP) and Com-plex Proportional Assessment Method (COPRAS) to show the importance of MOOC platforms and their impact in smart education.

E-learning is a very dynamic field, constantly growing, which refers to an educa-tional content or learning experiences provided or made by means of digital tech-nologies. The development of this area has a direct impact on the quality of edu-cation and cost reduction. E-learning is now dominated by learning management systems such as Moo-dle, to improve the interoperability of these systems; the use of data / resources is a major requirement, this need to push the e-Learning community to seek new approaches to improve this interoperability of systems. On the other hand, se-mantic Web technologies offers standards like RDF allows to describe Web re-sources with their metadata. We propose in this work an approach based on Se-mantic Web technology to improve the interoperability of e-Learning systems like Moodle.

The evolution of mobile learning has changed the way of learning. With the growing popularity of using mobile devices, mobile learning has become important in teaching and learning in education. Most m-learning services are provided by Learning Management Systems (LMS). This article aims at designing a software architecture that focuses in particular on the integration of SCORM's next-generation specifications in order to adapt the features and mechanisms of the learning management system and its extension to the mobile scenario. the architecture has brought several interactivities that are not feasible with the specifications of the next generation SCORM. This architecture is implemented for the famous open source LMS Moodle. The document is intended for developers and stakeholders of Mobile Learning, as well as educators and technologists addressing standards and interoperability specifications learning in general.