Athenea Engineering sciences journal

Modeling and Simulation of Intelligent Systems for the Early Detection of Academic Stress in University Students Using Machine Learning

2026-06-27T22:24:01+00:00

This study develops an intelligent system based on machine learning for the early detection of academic stress in university students, focusing on the areas of Engineering and Social Sciences. Based on a quantitative and simulation-based approach, a structured dataset was built integrating academic, behavioral, and psychosocial variables. Predictive models were implemented, including Random Forest, Support Vector Machines, and XGBoost, and their performance was evaluated using multiclass and binary classification metrics. The results show that ensemble models achieved the highest levels of accuracy, exceeding 90% in the detection of stress levels. Likewise, key variables such as fatigue, concentration, pending assignments, and hours of sleep were identified. Scenario simulation demonstrated that combined interventions generate significant reductions in the probability of high stress. The study provides a predictive-explanatory approach that contributes to decision-making in educational environments oriented toward student well-being.

Computational Models for Reducing Cognitive Overload in Business Administration Students

2026-06-27T22:23:52+00:00

The study developed a predictive computational model to identify and reduce cognitive overload in undergraduate business administration students in digital environments. A quantitative approach was applied using academic, cognitive, and technological variables related to digital fatigue, academic stress, and the use of virtual platforms. Machine learning algorithms such as logistic regression, SVM, and Random Forest were implemented and complemented with computational simulation. The results showed that Random Forest achieved the best predictive performance and made it possible to detect critical scenarios of cognitive saturation. It is concluded that computational models can contribute to the design of intelligent educational systems oriented toward academic well-being.

Design of an Artificial Intelligence-Based Platform for the Adaptive Management of Anxiety in Digital University Environments

2026-06-27T22:23:43+00:00

The digital transformation of higher education has increased the need to develop technological strategies and engineering designs aimed at supporting students' emotional well-being in virtual environments. The objective of this study was to develop an engineering design to support anxiety management in digital university environments. The study adopted a quantitative approach with a quasi-experimental design and involved 240 university students distributed into an experimental group and a control group. The platform integrated emotional monitoring, predictive analysis, and adaptive intervention mechanisms through machine learning algorithms. The results showed a significant reduction in anxiety, stress, and cognitive load in the experimental group, as well as high levels of predictive accuracy in the implemented models. It is concluded that adaptive artificial intelligence represents a viable alternative for strengthening more human-centered and preventive digital educational ecosystems.

Artificial Intelligence Model for Mitigating Academic Stress in Management Students

2026-06-27T22:23:34+00:00

Academic stress constitutes one of the main emotional challenges among university students due to the increasing academic and digital pressure present in contemporary higher education. This study aimed to develop an artificial intelligence model for mitigating academic stress in management students through machine learning techniques and predictive simulation. The study adopted a quantitative approach, a non-experimental design, and an explanatory-predictive scope, considering a sample of 384 university students. Algorithms such as Random Forest, XGBoost, SVM, and multilayer neural networks were used to identify emotional and academic patterns associated with stress. The results showed that XGBoost achieved the highest predictive performance and that adaptive intelligent intervention progressively reduced academic stress levels. It is concluded that artificial intelligence has strong potential to strengthen preventive strategies and emotional support systems in higher education.

Artificial Intelligence-Generated Laboratory Protocols for Chemistry Education: Safety, Didactic Value, and Green Chemistry

2026-06-29T22:28:32+00:00

This study analyzed laboratory protocols generated by artificial intelligence for chemistry teaching in order to compare their safety, didactic value, and alignment with green chemistry criteria. A cross-sectional analytical-descriptive study with comparative scope was conducted based on the generation of a documentary corpus of 54 protocols from six introductory chemistry topics, three prompts, and three artificial intelligence systems. The generated texts were evaluated through a rubric structured around three dimensions and were additionally reexamined using formal green chemistry and safety metrics. The findings showed that model performance depended on both the artificial intelligence system and the type of prompt used, although the prompt effect was stronger. Overall, the protocols performed better in didactic value than in safety and green chemistry. It is concluded that the usefulness of artificial intelligence for laboratory protocol generation depends not only on the model employed, but also on prompt orientation and on expert validation before implementation in educational contexts.

Nonlinear Dynamic Modeling and Monte Carlo Simulation of Failure Propagation in Antifragile Energy Networks under Stochastic Perturbations

2026-06-29T22:28:23+00:00

Modern energy networks constitute complex dynamic systems characterized by operational uncertainty and nonlinear behavior. The objective of this research was to develop a physical-computational framework to analyze the stability and adaptive capacity of energy networks subjected to stochastic perturbations. A nonlinear dynamic model based on ordinary differential equations was employed, integrating Monte Carlo simulation, Latin Hypercube sampling, an Energy Antifragility Index (EAI), sensitivity analysis using Sobol indices, and bifurcation analysis. The results revealed fragile, resilient, and antifragile behaviors, with resilient scenarios predominating. Coupling intensity and perturbation magnitude were the parameters with the greatest influence on the system. Likewise, a critical threshold associated with the emergence of multiple equilibrium states and dynamic transitions was identified. It is concluded that the integration of nonlinear dynamics and probabilistic simulation makes it possible to understand the behavior of complex energy systems under uncertainty.

Predictive Modeling Based on Artificial Intelligence and Smart Sensors for the Dynamic Prevention of Occupational Accidents in High-Risk Industrial Environments

2026-06-30T22:31:54+00:00

The prevention of occupational accidents in high-risk industrial environments requires solutions capable of anticipating unsafe conditions in real time. The objective of this research was to develop a predictive model based on artificial intelligence and smart sensors for the dynamic prevention of occupational accidents. The methodology integrated Industrial Internet of Things (IIoT) technologies, machine learning algorithms, Monte Carlo simulation, and explainable artificial intelligence to analyze environmental, operational, and human variables. The results showed that the XGBoost model achieved an accuracy of 94% and an AUC of 0.97, while the SHAP analysis identified operational fatigue, abnormal vibrations, and ambient temperature as the main risk-related factors. It is concluded that the integration of smart sensors, artificial intelligence, and advanced simulation strengthens industrial safety through predictive systems aimed at early risk detection and preventive decision-making.

Multi-Robot Platform for Remote Robotics Teaching in Virtual Environments

2026-06-30T23:38:46+00:00

This study evaluates the technical performance and pedagogical impact of a remote multi-robot platform based on a centralized architecture for teaching IoT and artificial vision. The technical feasibility of the platform was validated through a quality of service (QoS) analysis in different residential connectivity environments, showing stable operation with adequate levels of latency, \textit{jitter}, and packet loss. A quasi-experimental design was implemented with 56 students distributed into two groups: ESPE, which used the multi-robot platform, and ISUCT, which used \textit{software}-based simulators. The results showed significant differences in favor of ESPE in overall academic performance, as well as higher levels of student satisfaction. Additionally, IoT showed better results than artificial vision in both institutions due to its lower cognitive complexity. The findings support the use of remote laboratories based on cyber-physical systems as an effective alternative to strengthen practical training.