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.

Currently, massive open online courses (MOOCs) are experiencing major developments and arebecoming increasingly popular in distance learning programs. The goal is to break down inequalitiesand disseminate knowledge to everyone by creating a space for exchange and interaction.Despite the improvements to this educational model, MOOCs still have low retention rates, whichcan be attributed to a variety of factors, including learners’ heterogeneity. The paper aims toaddress the issue of low retention rates in MOOCs by introducing an innovative prediction modelthat provides the best (optimal) learning path for at-risk learners. For this purpose, learners at riskof dropping out are identified, and their courses are adapted to meet their needs and skills. A casestudy is presented to validate the effectiveness of our approach using classification algorithms forprediction and the ant colony optimization (ACO) algorithm to optimize learners’ paths.

The emphasis of recent works is focused on the use of rules based systems that describe the system behavior by simple rules which increase the flexibility and easiness of programming. In the context of the distributed test where a set of parallel testers exchange 110 messages to perform the test, some potential problems of coordination can arise amongst remote testers. These problems are usually known as controllability and Observability issues. We present in this paper our approach based on the use of a rule based multi-agent system to overcome such problems.

The paper presents some technical issues for testing distributed systems. The proposed approach consists on exploring the temporal properties that specify the time required to exchange messages between the various components of the distributed test application. Thereby taking into consideration the temporal properties in the specification of the behavior of conformance testing, it will provide a higher degree of expressiveness and flexibility. In other side, the paper show how to cope with problems encountered in distributed testing using timed distributed testing rules. Afterwards, we present our algorithm allowing the generation of such rules and a prototype of test using a Distributed Algorithmic and Rule-Based System (DARBS).