Background/Objectives: Cardiovascular diseases (CVDs) remain the primary cause of mortality worldwide, underscoring the critical importance of developing accurate early prediction models. In this study, we propose an advanced stacking ensemble learning framework to improve the predictive performance for CVD diagnosis. Methods: The methodology encompasses comprehensive data preprocessing, feature selection, cross-validation, and the construction of a stacking architecture integrating Random Forest (RF), Support Vector Machine (SVM), and CatBoost as base learners. Two meta-learning configurations were examined: Logistic Regression (LR) and a Multilayer Perceptron (MLP). Results: Experimental results indicate that the MLP-based stacking model achieves superior performance, with an accuracy of 97.06%, outperforming existing approaches reported in the literature. Furthermore, the model demonstrates high recall (96.08%) and precision (98%), confirming its robustness and generalization capacity. Conclusions: Compared to individual classifiers and traditional ensemble methods, the proposed approach yields significantly enhanced predictive outcomes, highlighting the potential of deep learning-based stacking strategies in cardiovascular risk assessment.

Artificial intelligence (AI) is increasingly impacting the medical field, especially in nutrition, where machine learning algorithms offer new opportunities for personalized dietary assessment. This paper presents a prototype of an AI-based expert system designed to support meal planning and nutrient intake evaluation, addressing individual dietary requirements, preferences, and constraints. The system employs essential image processing steps such as resizing, normalization, and magnification to ensure the accuracy and consistency of food image data. Using advanced image recognition models, the system processes visual food inputs to classify dishes and support nutritional analysis. The FOOD-MOR Dataset, comprising diverse traditional and modern Moroccan dishes, serves as the foundation for training and testing the models. A comparative evaluation of two pre-trained convolutional neural networks, VGG16 and MobileNetV2, was conducted. Using VGG16, the model achieved 81% accuracy with a macro F1-score of 0.80, while MobileNetV2 reached 86% accuracy and a 0.86 macro/weighted F1-score, demonstrating superior performance on the 10-class food classification task. These results indicate that the lighter MobileNetV2 architecture is more suitable for efficient, realtime food recognition. This study highlights the potential of AI-powered food recognition systems to enhance personalized nutrition by accurately identifying meals and facilitating dietary monitoring in diverse culinary contexts and cuisines.