The ability of Generative Adversarial Networks (GANs) to produce images that closely resemble real ones has raised concern. This requires the creation of efficient detection techniques because it has significant ramifications for digital media, security, and ethics. In order to demonstrate the growing difficulties of attaining authenticity in the rapidly developing field of Artificial Intelligence (AI), this study introduces this critical issue by leveraging the “Detect AI-Generated Faces: High-Quality Dataset,” obtained from Kaggle which contains 3,203 images of real human faces and AI-generated faces. However, the Orange3 data mining framework is used to analyze these images, focusing on extracting essential features such as shape attributes, texture descriptors, and color histograms. The dataset was divided into a training set (70%) and a testing set (30%) to evaluate our models effectively. Also, four machine learning algorithms were employed: K-Nearest Neighbor (KNN), Artificial Neural Network (ANN), Adaptive Boosting (AdaBoost), and Gradient Boosting (GB). The results revealed that KNN and AdaBoost achieved impressive accuracies of 99.4% and 97.07%, respectively, while GB and ANN reached even higher accuracies of 99.8% and 99.9%. These results underscore the effectiveness of advanced machine learning techniques in accurately distinguishing between AI-generated and real faces.
Areen Arabiat· International Journal of Ele...· 16 citations
This literature review examines the opportunities, implementation challenges, ethical implications, and emerging directions of predictive learning analytics (PLA) in engineering education. Using a structured review of the literature, the study synthesizes evidence from several publications with emphasis on studies examining risk prediction, personalized support, curricular improvement, interpretability, fairness, and intervention design. The review shows that PLA can improve early identification of at-risk students, support adaptive learning pathways, and inform data-driven refinements in engineering curricula; however, its impact depends on data quality, model transparency, institutional capacity, and the availability of timely human support. The analysis further indicates that the most consequential barriers are fragmented data ecosystems, the difficulty of translating predictions into effective interventions, and unresolved ethical concerns related to privacy, bias, consent, and student agency. The article contributes to educational research by offering an integrated synthesis that connects technical development with pedagogical evaluation and ethical governance in engineering education. It concludes by proposing that future PLA adoption should align predictive modeling with explainable artificial intelligence, learning-theory-informed intervention design, and institution-level implementation strategies. Publications were selected for relevance to PLA in engineering education and then synthesized narratively across opportunities, challenges, ethics, and future directions.
Hamza Abu Owida, Areen Arabiat· International Journal of Eva...· 0 citations
This study emphasizes the potential of computational techniques in cancer risk assessment, highlighting opportunities for specific and data-driven healthcare solutions. It examines the use of artificial intelligence (AI), machine learning (ML), and deep learning (DL) approaches to improve cancer risk assessment using a Kaggle dataset. The study uses Java-based ML software to create and evaluate multiple predictive models, taking advantage of its powerful libraries and frameworks for processing and analyzing cancer risk indicators. This work analyzes model performance using 10-fold cross-validation, resulting in reliable generalization and accuracy estimates. Several classification techniques, such as Random Forest (RF), logistic regression (LR), decision trees (DT), Naive Bayes (NB), and Multi-layer perceptron (MLP), are used to assess their efficacy in predicting risk levels for various cancer types. To measure classification effectiveness, key performance metrics such as accuracy, precision, recall, and F1 score are produced, in addition to multi-class confusion matrices. The results show that the RF model is the best classifier for classification, with accuracy of 99.85%, F-measure of 99.80%, precision of 99.80%, and sensitivity of 99.90%. These findings demonstrate the model’s ability to effectively estimate cancer risk levels among individuals, allowing for earlier discovery and more effective medical care.
Areen Arabiat, H. Owida, Suhaila Abuowaida et al.· bioRxiv· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
