Background: There is an immense debate about the factors that could limit the adoption of agroforestry systems. However, one of the most important is the generation of scientific information that supports the viability and benefits of the proposed techniques. Statistical analysis: This work used the Latent Dirichlet Allocation (LDA) modeling method to identify and interpret scientific information on topics in relation to existing categories in a set of documents. It also used the HJ-Biplot method to determine the relationship between the analyzed topics, taking into consideration the years under study. Results: A review of the literature was conducted in this study and a total of 9794 abstracts of scientific articles published between 1993 and 2022 were obtained. The United States, India, Brazil, the United Kingdom, and Germany were the five countries that published the largest number of studies about agroforestry, particularly soil organic carbon, which was the most studied case. The five more frequently studied topics were: soil organic carbon, adoption of agroforestry practices, biodiversity, climatic change global policies, and carbon and climatic change. Conclusion: the LDA and HJ-Biplot statistical methods are useful tools for determining topicality in text analysis in agroforestry and related topics.

The circular economy (CE) has been adopted as an alternative to the linear economic model, since it transforms the way products are designed and consumed. Agri-food chains are key to ensuring the transition to a sustainable development model in line with the principles of CE. This study sought to implement the concepts of circular economy at the Ecuador's banana production chain, through the design of prospective from the exhaustive diagnosis of the main processes of the chain. The research was conducted from 2020 to 2022. The studied checklist included 91 items grouped into nine dimensions: source or supply of materials, design, manufacturing, economic circle, distribution and sales, consumption and use, 4R, re-manufacturing, and sustainability. A medium circular economy level (CEL) was evaluated (2.69 points out of 5). Two machine learning techniques (MLTs) were also applied support vector machine (SVM), (with three types of kernel: radial, linear and polynomial) and a neural network (NN). In relation to the Spearman correlation coefficients of the application of the techniques to the study cases, they were higher than 0.90. The most reliable was the SVM Regression-Linear Kernel, which had a coefficient close to 1. Finally, various value-adding options are proposed in order to take advantage of the large amount of waste generated in the banana production system. The proposed options include value-augmentation processes, ranging from organic fertilizers, biofuels, materials for wastewater treatment, and the production of bioplastics and nanoparticles, which can be implemented using the CE approach.

The objective of this paper is derived from the theoretical analysis of the application of support vector machines to the design and management of agri-food chains. This analysis is conducted with an empirical approach, for the prediction of the level of integration in agri-food chains through support vector machines. The methodology designed and used for the processing of research results, which consists in the training of support vector machines is used to predict the level of integration in an agri-food chain. This type of predictive application appears in the literature consulted on the integration of agri-food chains. The analysis is performed comparing the method proposed with the neural network technique. The results of this research are mainly focused on predicting the level of integration in agri-food chains through vector machines. The study provides a support vector machine model that is applied to other case studies and therefore, allows predicting the outcome. The paper also shows the comparison of two techniques that share the goal of predicting, as applied in different contexts.

Este estudio investiga la aplicación de técnicas de Machine Learning (ML), específicamente el algoritmo de clustering K-Means y la reducción de dimensionalidad mediante PCA, para identificar patrones en datos de siniestros de tránsito. Analizando un conjunto de datos que incluye 21,352 registros de siniestros, este trabajo clasifica los accidentes en cuatro clusters distintos, revelando diferencias significativas en las características de los siniestros relacionadas con la ubicación, hora del día, causas y condiciones ambientales. Los resultados destacan la complejidad de los factores que contribuyen a los siniestros de tránsito y subrayan la necesidad de intervenciones personalizadas para mejorar la seguridad vial. Este enfoque permite una comprensión más profunda de los patrones de siniestros, facilitando el desarrollo de políticas y estrategias de prevención más efectivas. Este estudio no solo amplía la literatura existente, ofreciendo una metodología analítica robusta para el análisis de datos de siniestros, sino que también proporciona orientación práctica para la implementación de intervenciones de seguridad vial dirigidas y basadas en evidencia. Recomendamos futuras investigaciones para incluir más variables, comparar diferentes modelos de ML y realizar estudios longitudinales que permitan evaluar la evolución de los patrones de siniestros y la efectividad de las políticas de seguridad vial implementadas.

El objetivo de este trabajo es predecir el nivel de desarrollo de una cadena de suministro manabita de clasificación agroalimentaria, cuyo producto fundamental es el agua de coco artesanal. Usando datos históricos, se aplican los métodos de Máquinas de Soporte Vectorial en Regresión y Regresión Lineal Múltiple. Como parte del trabajo se lleva a cabo un estudio de correlación para mejorar el desempeño de los métodos. Los resultados muestran que las Máquinas de Soporte Vectorial logran el mejor desempeño cuando se usa el Kernel Polinomial con una selección de parámetros apropiados. La metodología empleada para el procesamiento de la investigación utiliza el entrenamiento de estos dos métodos lo que proporciona un modelo que se puede aplicar en otros estudios de casos.