Teacher job satisfaction is an important aspect of academic performance, student retention, and teacher retention. We propose to determine the predictive model of job satisfaction of basic education teachers using machine learning techniques. The original data set consisted of 15,087 instances and 942 attributes from the national survey of teachers from public and private educational institutions of regular basic education (ENDO-2018) carried out by the Ministry of Education of Peru. We used the ANOVA F-test filter and the Chi-Square filter as feature selection techniques. In the modeling phase, the logistic regression algorithms, Gradient Boosting, Random Forest, XGBoost and Decision Trees-CART were used. Among the algorithms evaluated, XGBoost and Random Forest stand out, obtaining similar results in 4 of the 8 metrics evaluated, these are: balanced accuracy of 74%, sensitivity of 74%, F1-Score of 0.48 and negative predictive value of 0.94. However, in terms of the area under the ROC curve, XGBoost scores 0.83, while Random Forest scores 0.82. These algorithms also obtain the highest true-positive values (479 instances) and lowest false-negative values (168 instances) in the confusion matrix. Economic income, satisfaction with life, self-esteem, teaching activity, relationship with the director, perception of living conditions, family relationships; health problems related to depression and satisfaction with the relationship with colleagues turned out to be the most important predictors of job satisfaction in basic education teachers.

Amazon rainforests are critical for providing a wide range of ecosystem services. In the Southeastern Peruvian Amazon; however, goldmining activities are causing severe soil degradation and forest loss. We analyzed aboveground biomass (AGB), forest structure, and species diversity recovery during secondary succession in 179 forest plots. Our study provides the first field-based quantification of AGB recovery following the abandonment by two types of goldmining (heavy machinery and suction pumping) in Madre de Dios (Peru). We found that successional secondary forests in areas subjected to suction pumping were more resilient than those in areas subjected to heavy machinery. After 20 years, mean AGB in suction pumping mining areas had reached 56% of reference forest AGB, while in areas of heavy machinery mining it was only 18%. Mining type, stand age, and distance from the forest edge had a significant effect on AGB. The influence of the distance from the forest edge on AGB varies according to mining type because the effects of species diversity on AGB are mediated by the distance from the forest edge. Our results clearly showed the dynamics of AGB recovery across a secondary succession after goldmining, and the contrasting responses of AGB between the two mining types. Our study disentangles the importance of key factors in forest recovery after mining and improves understanding of the resilience of biomass accumulation in these highly degraded ecosystems.

Teacher life satisfaction is crucial for their well-being and the educational success of their students, both essential elements for sustainable development. This study identifies the most relevant predictors of life satisfaction among Peruvian teachers using machine learning. We analyzed data from the National Survey of Teachers of Public Basic Education Institutions (ENDO-2020) conducted by the Ministry of Education of Peru, using filtering methods (mutual information, analysis of variance, chi-square, and Spearman’s correlation coefficient) along with embedded methods (Classification and Regression Trees—CART; Random Forest; Gradient Boosting; XGBoost; LightGBM; and CatBoost). Subsequently, we generated machine learning models with Random Forest; XGBoost; Gradient Boosting; Decision Trees—CART; CatBoost; LightGBM; Support Vector Machine; and Multilayer Perceptron. The results reveal that the main predictors of life satisfaction are satisfaction with health, employment in an educational institution, the living conditions that can be provided for their family, and conditions for performing their teaching duties, as well as age, the degree of confidence in the Ministry of Education and the Local Management Unit (UGEL), participation in continuous training programs, reflection on the outcomes of their teaching practice, work–life balance, and the number of hours dedicated to lesson preparation and administrative tasks. Among the algorithms used, LightGBM and Random Forest achieved the best results in terms of accuracy (0.68), precision (0.55), F1-Score (0.55), Cohen’s kappa (0.42), and Jaccard Score (0.41) for LightGBM, and accuracy (0.67), precision (0.54), F1-Score (0.55), Cohen’s kappa (0.41), and Jaccard Score (0.41). These results have important implications for educational management and public policy implementation. By identifying dissatisfied teachers, strategies can be developed to improve their well-being and, consequently, the quality of education, contributing to the sustainability of the educational system. Algorithms such as LightGBM and Random Forest can be valuable tools for educational management, enabling the identification of areas for improvement and optimizing decision-making.

Liquefied petroleum gas (LPG), composed mainly of propane and butane, is widely used as an energy source in residential, commercial, and industrial sectors; however, its high flammability poses a critical risk in the event of accidental leaks. In Peru, where LPG constitutes the main domestic energy source, leakage emergencies affect thousands of households each year. This pattern is replicated in developing countries with limited energy infrastructure. Early quantitative detection of propane, the predominant component of Peruvian LPG (~60%), is essential to prevent explosions, poisoning, and greenhouse gas emissions that hinder climate change mitigation strategies. This study presents PropNet-R, a convolutional neural network (CNN) designed to estimate propane concentrations (ppm) from thermal images. A dataset of 3574 thermal images synchronized with concentration measurements was collected under controlled conditions. PropNet-R, composed of four progressive convolutional blocks, was compared with SqueezeNet, VGG19, and ResNet50, all fine-tuned for regression tasks. On the test set, PropNet-R achieved MSE = 0.240, R² = 0.614, MAE = 0.333, and Pearson’s r = 0.786, outperforming SqueezeNet (MSE = 0.374, R² = 0.397), VGG19 (MSE = 0.447, R² = 0.280), and ResNet50 (MSE = 0.474, R² = 0.236). These findings provide empirical evidence that task-specific CNN architectures outperform generic transfer learning models in thermal image-based regression. By enabling continuous and quantitative monitoring of gas leaks, PropNet-R enhances safety in industrial and urban environments, complementing conventional chemical sensors. The proposed model contributes to the development of sustainable infrastructure by reducing gas-related risks, promoting energy security, and strengthening resilient, safe, and environmentally responsible urban systems.

La cuantificación de parámetros fisicoquímicos y microbiológicos es fundamental para definir la calidad del agua, pero el entendimiento de las relaciones y el comportamiento de estos parámetros, contribuyen a definir el estado trófico y calidad ambiental de los ecosistemas acuáticos. En este contexto, el objetivo del estudio fue determinar las relaciones entre los parámetros fisicoquímicos y microbiológicos del agua de la bahía interior del lago Titicaca en la ciudad de Puno-Perú, aplicando la técnica de árboles de predicción de aprendizaje automático. La muestra estuvo constituida por 17 parámetros de calidad, que fueron medidos mensualmente durante enero de 2013 y abril de 2019, obteniéndose un total de 76 datos por cada parámetro. Para la elaboración de los árboles de predicción y su posterior interpretación, se utilizó el paquete estadístico R. Se evidenció que los parámetros fisicoquímicos preponderantes del lago Titicaca, son temperatura, oxígeno disuelto, potencial de hidrógeno, conductividad eléctrica, sólidos suspendidos totales, clorofila-a, demanda bioquímica de oxígeno, demanda química de oxígeno y transparencia, a partir de los cuales se aplicó la técnica de árboles de predicción y se comprobó la eficiencia de estos modelos estadístico en la predicción de dichos parámetros para realizar estimaciones, teniendo en cuenta las características del lugar donde se utilice.