College dropout represents a significant challenge for universities, and despite advances in machine learning technologies, predicting dropout remains a complex task. This literature review focuses on investigating the factors that influence college dropout, examining the models used to predict it, and highlighting the most significant advances in explainability and simulation over the period 2012 to 2024 using the PRISMA methodology. They identified 520 factors in five categories (demographic, socioeconomic, institutional, personal, and academic), with the most studied factors in each category being, respectively, gender, scholarships, infrastructure, student identification, and grades. They also identified 83 machine learning models, with the most studied being the decision tree, logistic regression, and random forest. In addition, eight explanatory models were identified, with SHAP and LIME being the most widely used. Finally, no simulation models related to university dropout were identified. This study groups factors related to university dropout into key models for prediction and analyzes the methods used to explain the causal factors that influence university student dropout.

The teaching–learning process, at each educational level, is often an open problem for educators and researchers related to the stated topic. Researchers combine emerging technologies to formulate learning tools in order to understand the abstract contents of the subjects; however, the problem still persists. A technological learning tool would be effective when projected into an educational model that looks at motivation, usability, engagement, and technological acceptability. Some of these aspects could be attributed through the use of augmented reality and games. The aim of this work is to analyze, in the literature, the trends of learning models based on computer technologies for an effective and enjoyable learning activity. The analysis of the literature in that context—emphasizing acceptability, categories, entertainment, educational models—shows that it is still not well explored.

Health insurance fraud (HIF) places a substantial economic burden on global health systems. While supervised machine learning (SML) offers a promising solution for its detection, most approaches are ad hoc and lack a systematic methodological framework that ensures replicability, adaptability, and effectiveness, especially in contexts with severe class imbalance. We developed PDHIF (Phases for Detecting Fraud in Health Insurance), a six-phase systematic methodology that introduces a holistic focus that integrates fraud theory, actors, manifestations, and factors with the complete SML lifecycle. We applied this methodology in a case study using a dataset of 8.5 million claims from a public health insurance system in Peru. We trained and evaluated three SML models (Random Forest, XGBoost, and multilayer perceptron) in two experimental scenarios: one with the original, highly unbalanced dataset and another with a training set balanced via the K-means SMOTE technique. When PDHIF was applied, the results revealed a stark contrast: in the unbalanced scenario, the models were ineffective at detecting fraud (F1 score < 0.521) despite high accuracy (>98%). In the balanced scenario, the performance improved dramatically. The best-performing model, RF, achieved an F1 score of 0.994, a sensitivity of 0.994, and an AUC of 0.994 on the test set, demonstrating a robust ability to distinguish suspicious claims.

Objectives: Advantages of Information and Communication Technologies (ICT) in the educational process objectify us to explore the advances of Learning Management Systems (LMS), and preferences of Social Network Sites (SNS), as effective and attractive platforms in university education. Method/Analysis: Pursuing the objective, a set of articles related to the topic are analyzed qualitatively, according to the characteristics of the LMS tools in the improvement of the interaction in the educational process, as well as the characteristics of the SNS for virtual learning communities, and the dimensions that the interaction of these two tools must acquire, for an effective teaching-learning process. Findings: The work shows the trends of use of SNS, in particular Facebook, in informal learning that can be used efficiently in formal learning by attaching to LES platforms, in particular Moodle, with pedagogical, motivational and construction rules Social. Diverse discussions in the literature show that the LMS are not fulfilling the objectives proposed because they lack a social presence, an effective motivation to use, and because they tend to be tendencies to be technical repositories. Applications/Improvements: As a consequence, we suggest models of educational focus based on the synergies of LMS tools and social networks with technological, pedagogical, cognitive and social approaches for face-to-face and virtual education.