Jesús Alejandro Navarro-Acosta

Prof. Jesús Alejandro Navarro-Acosta

UNIVERSIDAD AUTÓNOMA DE COAHUILA

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
24 July 2024 in Algorithms
Verified
Journal Article
Open access
PDF file
Jose Alberto Solis VillarrealJose Alberto Solis Villarreal
Jose Alberto Solis Villarreal
VSValeria Soto-Mendoza
Valeria Soto-Mendoza
Jesús Alejandro Navarro-AcostaJesús Alejandro Navarro-Acosta
Jesús Alejandro Navarro-Acosta
EREfrain Ruiz y Ruiz
Efrain Ruiz y Ruiz
The development of smart cities will require the construction of smart buildings. Smart buildings will demand the incorporation of elements for efficient monitoring and control of electrical consumption. The development of efficient AI algorithms is needed to generate more accurate electricity consumption predictions; therefore; anomaly detection in electricity consumption predictions has become an important research topic. This work focuses on the study of the detection of anomalies in domestic electrical consumption in Mexico. A predictive machine learning model of future electricity consumption was generated to evaluate various anomaly-detection techniques. Their effectiveness in identifying outliers was determined, and their performance was documented. A 30-day forecast of electrical consumption and an anomaly-detection model have been developed using isolation forest. Isolation forest successfully captured up to 75% of the anomalies. Finally, the Shapley values have been used to generate an explanation of the results of a model capable of detecting anomalous data for the Mexican context.
1

27 January 2022 in International Journal of Computer Integrated Manufacturing
Verified
Journal Article
Closed access
EREdgar O. Reséndiz-Flores
Edgar O. Reséndiz-Flores
Jesús Alejandro Navarro-AcostaJesús Alejandro Navarro-Acosta
Jesús Alejandro Navarro-Acosta
IGIrma D. García-Calvillo
Irma D. García-Calvillo
The development of a new methodology for the accurate fault detection considering only those really important features of practical impact is the main motivation of this work. Therefore, the Kernel-based Mahalanobis-Taguchi System as a pre-classification stage followed by a binary particle swarm optimization procedure is introduced in this work in order to identify variables of impact by means of a feature screening in an optimal manner for industrial manufacturing processes. In order to study the practical efficiency of the presented approach, a computational comparison between the introduced approach and the standard Mahalanobis-Taguchi System is reported. Furthermore to test the relevance of the selected features by the proposed fault diagnosis approach, the sequential feature selection (SFS) and Random Forest (RF) algorithms are implemented and five experiments were carried out training a total of 15 machine learning algorithms. The presented approach is applied to foam injection and welding processes from the manufacturing automotive industry showing that the presented approach can achieve betters accuracy values for fault detection by selecting a small subset of features than the other presented feature selection techniques.


21 December 2020 in Applied Sciences
Verified
Journal Article
Open access
PDF file
Jesús Alejandro Navarro-AcostaJesús Alejandro Navarro-Acosta
Jesús Alejandro Navarro-Acosta
Irma D. García-CalvilloIrma D. García-Calvillo
Irma D. García-Calvillo
Vanesa Avalos-GaytánVanesa Avalos-Gaytán
Vanesa Avalos-Gaytán
Edgar Reséndiz-FloresEdgar Reséndiz-Flores
Edgar Reséndiz-Flores
In this study, a system for faults detection using a combination of Support Vector Data Description (SVDD) with metaheuristic algorithms is presented. The presented approach is applied to a real industrial process where the set of measured faults is scarce. The original contribution in this work is the industrial context of application and the comparison of swarm intelligence algorithms to optimize the SVDD hyper-parameters. Four recent metaheuristics are compared hereby to solve the corresponding optimization problem in an efficient manner. These optimization techniques are then implemented for fault detection in a multivariate industrial process with non-balanced data. The obtained numerical results seem to be promising when the considered optimization techniques are combined with SVDD. In particular, the Spotted Hyena algorithm outperforms other metaheuristics reaching values of F1 score near 100% in fault detection.

1 October 2021 in Ingeniería Investigación y Tecnología
Verified
Journal Article
Open access
Jesús Alejandro Navarro-AcostaJesús Alejandro Navarro-Acosta
Jesús Alejandro Navarro-Acosta
VSValeria Soto-Mendoza
Valeria Soto-Mendoza
FZFélix Raymundo Saucedo Zendejo
Félix Raymundo Saucedo Zendejo
JEJosé Maria Guajardo Espinoza
José Maria Guajardo Espinoza
MMMaría Teresa Rivera Morales
María Teresa Rivera Morales
En la presente obra se describe la realización de un ejercicio de validación de resultados de una prueba psicológica aplicada a maestros y alumnos en estado de aislamiento por la pandemia por COVID-19 en el estado de Coahuila, México. El objetivo de este trabajo es aplicar técnicas de machine learning para validar un instrumento que mide las emociones y los sentimientos negativos, así como el sesgo cognitivo o desviación de pensamiento sobre la educación y la pandemia en situación de aislamiento. Para el cumplimiento del objetivo se aplicó un instrumento en formato electrónico que se diseminó en el estado de Coahuila, los usuarios responden y se genera la base de datos, la cual, después de su preprocesamiento es analizada mediante la combinación de Random forest (RF) y Support Vector Machines (SVM); obteniendo como resultado la pertinencia o no de algunos de los reactivos en las pruebas, dando con esto una validez interna al instrumento. Los resultados experimentales muestran que la metodología propuesta es capaz de seleccionar las variables predictoras más relevantes. De esta manera, se obtienen resultados satisfactorios en la clasificación y predicción de diagnósticos psicológicos globales y segmentados por características de los respondientes. Por otro lado, aunque las técnicas implementadas son robustas y confiables, éstas presentan limitaciones en cuanto a la observación de los otros tipos de validez: la de constructo, la externa, entre otras; lo cual pudiera limitar su utilización. Si bien, en el campo de la psicometría existen diversas estrategias clásicas, la metodología propuesta basada en la combinación de técnicas de machine learning para el análisis y validación de este tipo de pruebas, favorece el crecimiento de opciones para mejorar los diagnósticos y en consecuencia el tratamiento de padecimientos psicológicos.

16 October 2024 in Revista Mexicana de Economía y Finanzas
Verified
Journal Article
Open access
PDF file
Jesús Alejandro Navarro-AcostaJesús Alejandro Navarro-Acosta
Jesús Alejandro Navarro-Acosta
VSValeria Soto-Mendoza
Valeria Soto-Mendoza
LPLaura Policardo
Laura Policardo
ECEdgar Sanchez carrera
Edgar Sanchez carrera
Objective: The aim of this paper is to analyze the contrast of policies to face the Covid-19 pandemic in the socioeconomic performance of three representative economies: Italy, Mexico, and United States. Methodology: Machine learning (ML) techniques are applied to analyze the socioeconomic effects of the pandemic (containment measures, infection rates, total deaths, vaccination, etc.) on GDP growth in those countries. The experiment is that New Zealand's reference stringency index replaces each of those countries' own stringency index and the forecasts for GDP growth, Covid-19-induced deaths, and the Covid-19 reproduction rate. Thus, we show that ML techniques are robust tools for multiple outcome regressions and for experimental scenarios on the socioeconomic impact of the Covid-19 pandemic. Results: The experimental results revealed that the Regression Tree and Random Forest techniques successfully estimate and predict the cases of Italy, Mexico, and the United States. Conclusions: The proposal is that stringency measures and vaccination policies are undoubtedly successful in the fight against a pandemic, in addition to measuring the effects of such policies when data is available through the use of novel techniques such as ML.