Early detection of diabetes risk through supervised Machine Learning based on non-invasive individual factors

Authors

Jesús Ernesto González Torres
Departamento de Tecnologías de la Información y Comunicación. Universidad Estatal de Bolívar. Guaranda 020150, Ecuador. ROR: https://ror.org/005cgg117
https://orcid.org/0009-0007-0693-0320
Alexander Ufredo Alegria Chavez
Departamento de Tecnologías de la Información y Comunicación. Universidad Estatal de Bolívar. Guaranda 020150, Ecuador. ROR: https://ror.org/005cgg117
https://orcid.org/0009-0001-3700-2515
Diana Magali Alegría Camino
Carrera de Desarrollo de Software. Instituto Superior Tecnológico “El Libertador”. Guaranda 020150, Ecuador
https://orcid.org/0009-0002-3670-9479
Verónica Elizabeth Sánchez Aguiar
Carrera de Electrónica. Instituto Superior Tecnológico “El Libertador”. Guaranda 020150, Ecuador
https://orcid.org/0009-0003-4415-9566
DOI: https://doi.org/10.63804/mtc.v2i1.e5

Synopsis

The increasing incidence of chronic metabolic disorders, such as diabetes mellitus, represents a significant challenge to the sustainability of global healthcare systems. In response to the need for efficient and low-cost large-scale screening mechanisms, this study aimed to develop and evaluate an artificial intelligence model for the early detection of diabetes and prediabetes risk, relying exclusively on non-invasive individual factors to reduce dependence on complex clinical testing. A quantitative and experimental approach was applied, based on the training of supervised machine learning algorithms. The study processed a large dataset derived from behavioral surveillance systems, performing feature selection focused on physiological and lifestyle variables such as body mass index, age, and physical activity history. A gradient boosting architecture was implemented and optimized using an asymmetric class-weighting strategy designed to penalize false negatives in light of the inherent imbalance in medical data. The experimentation validated the robustness of the proposed technical approach. The model achieved sensitivity (recall) above 99 %, ensuring effective identification of the vast majority of at-risk individuals while minimizing the omission of critical cases. The integration of computational tools calibrated to maximize positive case detection constitutes a viable alternative to strengthen preventive medicine. It is concluded that this tool optimizes healthcare resource allocation by functioning as a reliable primary screening filter, enabling confirmatory clinical evaluations to be prioritized for patients with a higher real diagnostic probability.

Author Biographies

Jesús Ernesto González Torres, Departamento de Tecnologías de la Información y Comunicación. Universidad Estatal de Bolívar. Guaranda 020150, Ecuador. ROR: https://ror.org/005cgg117

Software Engineer graduated from the State University of Bolívar (UEB). He possesses solid professional experience in the technology industry, having held strategic roles such as Software Developer and Project Leader. Currently, he serves as a developer in the Department of Information and Communication Technologies (ICT) at UEB. His technical interests and research lines focus on advanced programming, database design and management, and the implementation of artificial intelligence solutions applied to real-world problems.

Alexander Ufredo Alegria Chavez, Departamento de Tecnologías de la Información y Comunicación. Universidad Estatal de Bolívar. Guaranda 020150, Ecuador. ROR: https://ror.org/005cgg117

Software Engineer holding a degree from the State University of Bolívar (UEB), Ecuador. He has professional experience as a software developer and currently serves as a Software Development Specialist in the Department of Information and Communication Technologies (ICT) at the same university. His research interests encompass programming, artificial intelligence, as well as blockchain technologies and cryptocurrencies.

Diana Magali Alegría Camino, Carrera de Desarrollo de Software. Instituto Superior Tecnológico “El Libertador”. Guaranda 020150, Ecuador

She holds a Licentiate degree in Education Sciences with a major in Educational Informatics and a Master’s degree in Information Systems, specializing in Business Intelligence and Big Data Analytics. Her background is complemented by the “Training of Trainers” certification, endorsed by the National System of Qualifications and Professional Training, which underpins her strong methodological competence. Currently, she serves as a professor in the Software Development program, where she integrates her dual pedagogical and technological profile. Her main areas of interest and research include programming, educational research, and the application of artificial intelligence, focusing on the use of emerging technologies for academic innovation and data analysis.

Verónica Elizabeth Sánchez Aguiar, Carrera de Electrónica. Instituto Superior Tecnológico “El Libertador”. Guaranda 020150, Ecuador

She holds a Licentiate degree in Education Sciences with a major in Educational Informatics and a Master’s degree in Information Systems, specializing in Business Intelligence and Big Data Analytics. Her methodological preparation is supported by the “Training of Trainers” certification from the National System of Qualifications and Professional Training. Currently, she serves as a professor in the Electronics program, where she applies her interdisciplinary background. Her primary areas of interest and professional development focus on Information Technologies, integrating data management tools and computer systems within the fields of engineering and technical education.

Cover MTC V2i1

Downloads

MTC V2i1 V 2026

Volume

Challenges of contemporary society

Pages

19

Published

February 28, 2026

Series

Market, Technology, and Citizenship

Online ISSN

3103-117X

Categories

Copyright (c) 2026 Thomás Ricardo Borja Saltos, Claudio Xavier Borja Saltos , Jorge Wilson Tamami Pachala, Cristhian Geovanny Villamarin Arroba, Mercedes Anabel Monar Verdezoto, Hernán Arturo Rojas Sánchez, Simeón Ángel Zavala Verdezoto, Alisson Daniela Ocampo Paño, Mario Manuel Paredes Lopez, Erika Anabel Duche Manobanda, Jesús Ernesto González Torres, Alexander Ufredo Alegria Chavez, Diana Magali Alegría Camino, Verónica Elizabeth Sánchez Aguiar (Autor de capítulo)

License

Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

How to Cite

Early detection of diabetes risk through supervised Machine Learning based on non-invasive individual factors. (2026). In Challenges of contemporary society: Vol. V2i1 (p. 19). Editorial Unión Científica. https://doi.org/10.63804/mtc.v2i1.e5