Julián Fernando Muñoz-Ordóñez

Prof. Julián Fernando Muñoz-Ordóñez

Systems Engineering, Corporación Universitaria Comfacauca - Unicomfacauca, Popayán, Cauca, Colombia
orcid

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

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

Representative Publications

View All Publications
11 May 2023 in Remote Sensing
Verified
Journal Article
Open access
PDF file
DADarwin Alexis Arrechea-Castillo
Darwin Alexis Arrechea-Castillo
YSYady Tatiana Solano-Correa
Yady Tatiana Solano-Correa
JMJulián Fernando Muñoz-Ordóñez
Julián Fernando Muñoz-Ordóñez
EPEdgar Pencue-Fierro
Edgar Pencue-Fierro
AFApolinar Figueroa-Casas
Apolinar Figueroa-Casas
Land Use and Land Cover (LULC) classification using remote sensing data is a challenging problem that has evolved with the update and launch of new satellites in orbit. As new satellites are launched with higher spatial and spectral resolution and shorter revisit times, LULC classification has evolved to take advantage of these improvements. However, these advancements also bring new challenges, such as the need for more sophisticated algorithms to process the increased volume and complexity of data. In recent years, deep learning techniques, such as convolutional neural networks (CNNs), have shown promising results in this area. Training deep learning models with complex architectures require cutting-edge hardware, which can be expensive and not accessible to everyone. In this study, a simple CNN based on the LeNet architecture is proposed to perform LULC classification over Sentinel-2 images. Simple CNNs such as LeNet require less computational resources compared to more-complex architectures. A total of 11 LULC classes were used for training and validating the model, which were then used for classifying the sub-basins. The analysis showed that the proposed CNN achieved an Overall Accuracy of 96.51% with a kappa coefficient of 0.962 in the validation data, outperforming traditional machine learning methods such as Random Forest, Support Vector Machine and Artificial Neural Networks, as well as state-of-the-art complex deep learning methods such as ResNet, DenseNet and EfficientNet. Moreover, despite being trained in over seven million images, it took five h to train, demonstrating that our simple CNN architecture is only effective but is also efficient.

1 January 2025 in International Journal of Intelligent Systems
Verified
Journal Article
Open access
JMJulián Fernando Muñoz-Ordóñez
Julián Fernando Muñoz-Ordóñez
Carlos CobosCarlos Cobos
Carlos Cobos
JVJuan Carlos Vidal
Juan Carlos Vidal
Francisco HerreraFrancisco Herrera
Francisco Herrera
The increasing adoption of artificial intelligence (AI) in critical domains such as healthcare, law, and defense demands robust mechanisms to ensure transparency and explainability in decision‐making processes. While machine learning and deep learning algorithms have advanced significantly, their growing complexity presents persistent interpretability challenges. Existing maturity frameworks, such as Capability Maturity Model Integration, fall short in addressing the distinct requirements of explainability in AI systems, particularly where ethical compliance and public trust are paramount. To address this gap, we propose the Maturity Model for eXplainable Artificial Intelligence: Analysis and Evaluation (MM4XAI‐AE), a domain‐agnostic maturity model tailored to assess and guide the practical deployment of explainability in AI‐based applications. The model integrates two complementary components: an analysis model and an evaluation model, structured across four maturity levels—operational, justified, formalized, and managed. It evaluates explainability across three critical dimensions: technical foundations, structured design, and human‐centered explainability. MM4XAI‐AE is grounded in the PAG‐XAI framework, emphasizing the interrelated dimensions of practicality, auditability, and governance, thereby aligning with current reflections on responsible and trustworthy AI. The MM4XAI‐AE model is empirically validated through a structured evaluation of thirteen published AI applications from diverse sectors, analyzing their design and deployment practices. The results show a wide distribution across maturity levels, underscoring the model’s capacity to identify strengths, gaps, and actionable pathways for improving explainability. This work offers a structured and scalable framework to standardize explainability practices and supports researchers, developers, and policymakers in fostering more transparent, ethical, and trustworthy AI systems.

9 November 2018
Verified
Book Chapter
Closed access
PDF file
JMJulián Fernando Muñoz-Ordóñez
Julián Fernando Muñoz-Ordóñez
Carlos CobosCarlos Cobos
Carlos Cobos
MMMartha Mendoza
Martha Mendoza
EHEnrique Herrera-Viedma
Enrique Herrera-Viedma
FHFrancisco Herrera
Francisco Herrera
STSiham Tabik
Siham Tabik

26 June 2023 in INGENIERÍA Y COMPETITIVIDAD
Verified
Journal Article
Open access
PDF file
COCristian Camilo Muñoz Ordoñez
Cristian Camilo Muñoz Ordoñez
Carlos CobosCarlos Cobos
Carlos Cobos
JMJulián Fernando Muñoz-Ordóñez
Julián Fernando Muñoz-Ordóñez
Coffee is one of the most traded agricultural products internationally; in Colombia, it is the first non-mining-energy export product. In this context, the prediction of coffee crop yields is vital for the sector since it allows coffee growers to establish crop management strategies, maximizing their profits or reducing possible losses. This paper addresses crucial aspects of coffee crop yield prediction through a systematic literature review of documents consulted in Scopus, ACM, Taylor & Francis, and Nature. These documents were subjected to a filtering and evaluation process to answer five key questions: predictor variables used, target variable, techniques and algorithms employed, metrics to evaluate the quality of the prediction, and types of coffee reported. The results reveal different groups of predictor variables, including atmospheric, chemical, satellite-derived, fertilizer-related, soil, crop management, and shadow factors. The most recurrent target variable is yield, measured in bean weight per hectare or other measures, with one case considering leaf area. Predominant techniques for yield forecasting include linear regression, random forests, principal component analysis, cluster regression, neural networks, classification and regression trees, and extreme learning machines, among others. The most common metrics to evaluate the quality of predictive models include R², RMSE, MAE, MAPE, MRE, error deviation, Pearson's correlation coefficient, and standard deviation. Finally, robusta, arabica, racemosa, and zanguebariae are the most studied coffee varieties. El café es uno de los productos agrícolas más comercializados internacionalmente y en Colombia, es el primer producto de exportación no minero-energético. En este contexto, la predicción del rendimiento de los cultivos de café es vital para el sector, ya que permite a los caficultores establecer estrategias de manejo del cultivo, maximizando sus ganancias o reduciendo posibles pérdidas. En este artículo, se abordan aspectos cruciales de la predicción del rendimiento de los cultivos de café mediante una revisión sistemática de literatura de documentos consultados en Scopus, ACM, Taylor & Francis y Nature. Estos documentos se sometieron a un proceso de filtrado y evaluación para responder cinco preguntas clave: variables predictoras, variable objetivo, técnicas y algoritmos empleados, métricas para evaluar la calidad de la predicción y tipos de café reportados. Los resultados revelan distintos grupos de variables predictoras que incluyen factores atmosféricos, químicos, obtenidos vía satélite, relacionados con fertilizantes, suelo, manejo del cultivo y sombras. La variable objetivo más recurrente es el rendimiento medido en peso de granos por hectárea u otras medidas, con un caso que considera el área foliar. Entre las técnicas predominantes para predecir el rendimiento se encuentran la regresión lineal, los bosques aleatorizados, el análisis de componentes principales, la regresión por conglomerados, las redes neuronales, los árboles de clasificación y regresión y las máquinas de aprendizaje extremo, entre otras. Las métricas más comunes para evaluar la calidad de los modelos predictivos incluyen R², RMSE, MAE, MAPE, MRE, error estándar, coeficiente de correlación de Pearson y desviación estándar. Por último, las variedades de café más estudiadas son robusta, arábica, racemosa y zanguebariae.

13 December 2021 in Investigación e Innovación en Ingenierías
Verified
Journal Article
Open access
PDF file
VRVíctor Pinto Rodríguez
Víctor Pinto Rodríguez
JMJulián Fernando Muñoz-Ordóñez
Julián Fernando Muñoz-Ordóñez
MBMichael Pardo Burbano
Michael Pardo Burbano
Objetivo: Construir el primer prototipo de punto ecológico inteligente que asista al usuario en la correcta clasificación de residuos sólidos en punto de origen utilizando Deep Learning, aportando nuevas tecnologías que estén enfocadas en mitigar el impacto climático causado por la mala gestión de los residuos sólidos. Metodología: para el desarrollo de la investigación se implementó la metodología LEAN de Toyota y la herramienta Solid Works para diseñar un prototipo acorde a las necesidades de un punto ecológico, incluyendo un análisis estructural y pruebas de peso con el fin de medir su capacidad de resistencia. La metodología ágil aplicada para el desarrollo de la inteligencia artificial y la conformación del conjunto de imágenes fue programación extrema. Resultados y conclusiones: se construye un conjunto de datos de tres clases acorde a la resolución 2184 de 2019 de Colombia, que contiene un total de 400 muestras tomadas desde el prototipo. El modelo de clasificación utilizando una arquitectura de bajo coste computacional: MobileNet obtuvo una precisión del 97.91% en la validación y un coeficiente kappa de 0.95. El modelo entrenado se ejecuta sobre una Raspberry Pi con el fin de minimizar costos de producción. La investigación presenta una prueba de aceptación y usabilidad del prototipo.