Automatic prediction of growth and yield of legume plants using artificial intelligence models in a smart mobile application.
No Thumbnail Available
Date
2025-10-06
Journal Title
Journal ISSN
Volume Title
Publisher
Ecuador: La Maná: Universidad Técnica de Cotopaxi; Extensión La Maná, Carrera de Sistemas de Información
Abstract
This article presents the design, training, and validation of a smart mobile application for the automated prediction of growth and yield in legume plants, using supervised learning algorithms. Random forest and decision tree models were employed, trained on a multivariable dataset with 319 records and 67 quantitative and qualitative variables collected through IoT sensors and meteorological APIs. The decision tree model achieved a coefficient of determination of 0.76 for plant height and 0.87 for forage weight, surpassing the random forest model in accuracy, with R² values of 0.80 and 0.72, respectively. The application, developed in React Native and linked to a Django backend, allows the user to select the algorithm they wish to work with. Furthermore, the functional validation, carried out with 350 beneficiaries including farmers and students, showed a high level of acceptance: 91% positively rated the usability of the application, and 84% expressed intent for recurrent use. This proposal represents a significant contribution to the digital agriculture ecosystem, providing an accessible, accurate, and adaptable tool with potential for scalability and community adoption.
Description
Keywords
MACHINE LEARNING, SUPERVISED MODELS, SMART AGRICULTURE, DATASET
Citation
Soledispa Vera, H. S., Navarrete Cedeño, A. J., Borja, C. D., & Luna Murillo, R. A. (2025). Automatic prediction of growth and yield of legume plants using artificial intelligence models in a smart mobile application. Revista Ingenio Global, 4(2), 52–76. https://doi.org/10.62943/rig.v4n2.2025.320