Enfoque basado en algoritmos genéticos para mejorar la predicción de las complicaciones del infarto de miocardio optimización de la relación de desequilibrio entre clases
Introducción: Este artículo es el resultado de investigaciones realizadas en el marco de la tesis doctoral del programa de Doctorado en Ciencias de la Computación de la Universidad del Cauca en 2025. El infarto de miocardio (IM) es una afección de alta mortalidad que presenta complicaciones mortales en las primeras horas tras la hospitalización. La predicción temprana de estas complicaciones mejora el pronóstico y reduce los riesgos.
Problema: Los algoritmos de aprendizaje automático predicen posibles complicaciones; sin embargo, el desequilibrio de clases en los conjuntos de datos sesga las predicciones hacia la clase mayoritaria.
Objetivo: Mejorar el rendimiento predictivo de un modelo de aprendizaje automático reequilibrando la distribución de clases en el conjunto de datos de entrenamiento mediante un enfoque basado en datos.
Metodología: Se adaptó la metodología CRISP-DM.
Resultados: Proponemos un algoritmo genético (AG) para identificar la proporción óptima de reequilibrio a partir de los datos. El individuo con mejor rendimiento determinó la proporción óptima de reequilibrio mediante sobremuestreo aleatorio (ROS) utilizando un modelo XGBoost, que clasificó correctamente 7 de 8 casos (puntuación F1: 0,36), superando al modelo de referencia, que clasificó correctamente 6 de 8 casos (puntuación F1: 0,28).
Conclusiones: Un enfoque que optimiza la proporción de equilibrio de clases mejora la puntuación F1 y detecta una clase adicional en comparación con el modelo base sin equilibrio de clases.
Originalidad: Este artículo propone un mecanismo para determinar la proporción óptima con la que debe reequilibrarse un conjunto de datos desequilibrado.
Limitaciones: Se requiere procesamiento cuando se trabaja con conjuntos de datos más grandes.
Cómo citar
Licencia
Derechos de autor 2026 Ingeniería Solidaria
Esta obra está bajo una licencia internacional Creative Commons Atribución 4.0.
Compromiso ético y cesión de derechos
El autor debe declarar que su trabajo es original e inédito y que no se ha postulado a evaluación simultánea para su publicación por otro medio. Además, debe asegurar que no tiene impedimentos de ninguna naturaleza para la concesión de los derechos previstos en el contrato.
El autor se compromete a esperar el resultado de evaluación de la revista Ingeniería Solidaria, antes de considerar su presentación a otro medio; en caso de que la respuesta de publicación sea positiva, adicionalmente, se compromete a responder por cualquier acción de reivindicación, plagio u otra clase de reclamación que al respecto pudiera sobrevenir por parte de terceros.
Asimismo, debe declarar que, como autor o coautor, está de acuerdo por completo con los contenidos presentados en el trabajo y ceder todos los derechos patrimoniales, es decir, su reproducción, comunicación pública, distribución, divulgación, transformación, puesta a disposición y demás formas de utilización de la obra por cualquier medio o procedimiento, por el término de su protección legal y en todos los países del mundo, al Fondo Editorial de la Universidad Cooperativa de Colombia, de manera gratuita y sin contraprestación presente o futura.
[1] A. Kavita, P. Sonal, and T. D. L. Nguyen, “Comparative analysis of K-NN, naïve Bayes, and logistic regression for credit card fraud detection,” Revista Ingeniería Solidaria, vol. 19, no. 3, pp. 1–22, 2023, doi: [https://doi.org/10.16925/2357-6014.2023.03.05](https://doi.org/10.16925/2357-6014.2023.03.05?utm_source=chatgpt.com). DOI: https://doi.org/10.16925/2357-6014.2023.03.05
[2] H. Allam, L. I. Makubvure, B. K. Gyamfi, K. N. Graham, and K. Akinwolere, “Text classification: How machine learning is revolutionizing text categorization,” Information, vol. 16, no. 2, Art. no. 130, 2025, doi: [https://doi.org/10.3390/info16020130](https://doi.org/10.3390/info16020130?utm_source=chatgpt.com). DOI: https://doi.org/10.3390/info16020130
[3] M. Manoj Yadav and A.-H. Ronak, “Resampling imbalanced healthcare data for predictive modelling,” International Journal of Advanced Computer Science and Applications, vol. 16, no. 2, Art. no. 36, 2025, doi: [https://doi.org/10.14569/IJACSA.2025.0160204](https://doi.org/10.14569/IJACSA.2025.0160204?utm_source=chatgpt.com). DOI: https://doi.org/10.14569/IJACSA.2025.0160204
[4] X. Xiaoying, L. Huawen, Z. Shouzhen, L. Lingbin, and L. Wen, “A novel progressively undersampling method based on the density peaks sequence for imbalanced data,” Knowledge-Based Systems, vol. 216, Art. no. 106689, 2021, doi: [https://doi.org/10.1016/j.knosys.2020.106689](https://doi.org/10.1016/j.knosys.2020.106689?utm_source=chatgpt.com). DOI: https://doi.org/10.1016/j.knosys.2020.106689
[5] S. Tyagi and S. Mittal, “Sampling approaches for imbalanced data classification problem in machine learning,” in Proc. International Conference on Research in Intelligent Computing (ICRIC 2019), 2020, pp. 209–221, doi: [https://doi.org/10.1007/978-3-030-29407-6_17](https://doi.org/10.1007/978-3-030-29407-6_17?utm_source=chatgpt.com). DOI: https://doi.org/10.1007/978-3-030-29407-6_17
[6] D. Elreedy and A. F. Atiya, “A comprehensive analysis of Synthetic Minority Oversampling Technique (SMOTE) for handling class imbalance,” Information Sciences, vol. 505, pp. 32–64, 2019, doi: [https://doi.org/10.1016/j.ins.2019.07.070](https://doi.org/10.1016/j.ins.2019.07.070?utm_source=chatgpt.com). DOI: https://doi.org/10.1016/j.ins.2019.07.070
[7] A. Fernández, S. García, and F. Herrera, “SMOTE for learning from imbalanced data: Progress and challenges, marking the 15-year anniversary,” Journal of Artificial Intelligence Research, vol. 61, pp. 863–905, 2018, doi: [https://doi.org/10.1613/jair.1.11192](https://doi.org/10.1613/jair.1.11192?utm_source=chatgpt.com). DOI: https://doi.org/10.1613/jair.1.11192
[8] V. Rattan, R. Mittal, J. Singh, and V. M. (ND), “Analyzing the application of SMOTE on machine learning classifiers,” in Proc. International Conference on Emerging Smart Computing and Informatics (ESCI), 2021, pp. 692–695, doi: [https://doi.org/10.1109/ESCI50559.2021.9396962](https://doi.org/10.1109/ESCI50559.2021.9396962?utm_source=chatgpt.com). DOI: https://doi.org/10.1109/ESCI50559.2021.9396962
[9] M. Mukherjee and M. Khushi, “SMOTE-ENC: A novel SMOTE-based method to generate synthetic data for nominal and continuous features,” Applied System Innovation, vol. 4, no. 1, Art. no. 18, 2021, doi: [https://doi.org/10.3390/asi4010018](https://doi.org/10.3390/asi4010018?utm_source=chatgpt.com). DOI: https://doi.org/10.3390/asi4010018
[10] D. Ishani and P. Vibha, “A comparative study of SMOTE, Borderline-SMOTE, and ADASYN oversampling techniques using different classifiers,” in Proc. 3rd International Conference on Smart Data Intelligence (ICSMDI), Trichy, India, 2023, doi: [https://doi.org/10.1109/ICSMDI57622.2023.00060](https://doi.org/10.1109/ICSMDI57622.2023.00060?utm_source=chatgpt.com). DOI: https://doi.org/10.1109/ICSMDI57622.2023.00060
[11] G. Husain, D. Nasef, R. Jose, J. Mayer, and K. Bekbol, “SMOTE vs. SMOTEENN: A study on the performance of resampling algorithms for addressing class imbalance in regression models,” Algorithms, vol. 18, no. 1, Art. no. 37, 2025. (ND) DOI: https://doi.org/10.3390/a18010037
[12] M. Altalhan, A. Algarni, and M. T.-H. Alouane, “Imbalanced data problem in machine learning: A review,” IEEE Access, vol. 13, pp. 13686–13699, 2025, doi: [https://doi.org/10.1109/ACCESS.2025.3531662](https://doi.org/10.1109/ACCESS.2025.3531662?utm_source=chatgpt.com). DOI: https://doi.org/10.1109/ACCESS.2025.3531662
[13] J. C. Rella, G. Claeskens, R. Cao, and J. M. Vilar, “Instance-dependent cost-sensitive parametric learning,” Neurocomputing, vol. 615, Art. no. 128875, 2025, doi: [https://doi.org/10.1016/j.neucom.2024.128875](https://doi.org/10.1016/j.neucom.2024.128875?utm_source=chatgpt.com). DOI: https://doi.org/10.1016/j.neucom.2024.128875
[14] A. S. Desuky and S. Hussain, “An improved hybrid approach for handling class imbalance problem,” Arabian Journal for Science and Engineering, vol. 46, no. 4, pp. 3853–3864, 2021, doi: [https://doi.org/10.1007/s13369-021-05347-7](https://doi.org/10.1007/s13369-021-05347-7?utm_source=chatgpt.com). DOI: https://doi.org/10.1007/s13369-021-05347-7
[15] A. Newaz, M. S. Mohosheu, and M. A. Al Noman, “Predicting complications of myocardial infarction within several hours of hospitalization using data mining techniques,” Informatics in Medicine Unlocked, vol. 42, Art. no. 101361, 2023, doi: [https://doi.org/10.1016/j.imu.2023.101361](https://doi.org/10.1016/j.imu.2023.101361?utm_source=chatgpt.com). DOI: https://doi.org/10.1016/j.imu.2023.101361
[16] H. Shang, J. Wang, Y. Cong, H. Chen, and Y. Cui, “Risk prediction models for heart failure after myocardial infarction: A systematic review,” Medicine, vol. 104, no. 30, 2025, doi: [https://doi.org/10.1097/MD.0000000000043480](https://doi.org/10.1097/MD.0000000000043480?utm_source=chatgpt.com). DOI: https://doi.org/10.1097/MD.0000000000043480
[17] UCI Machine Learning Repository, “Complications of myocardial infarction: A database for testing recognition and prediction systems,” 2020. [Online]. Available: [https://archive.ics.uci.edu/](https://archive.ics.uci.edu/?utm_source=chatgpt.com) (SD)
[18] M. Pandey and S. S. Routaray, “Feature optimized hybrid model for prediction of myocardial infarction,” F1000Research, vol. 14, Art. no. 78, 2024, doi: [https://doi.org/10.12688/f1000research.160393.1](https://doi.org/10.12688/f1000research.160393.1?utm_source=chatgpt.com). DOI: https://doi.org/10.12688/f1000research.160393.1
[19] W. H. Elashmawi, A. Tamer, and A. Nasr, “Multi-target classification for predicting myocardial infarction complications: A comparative analysis of various machine learning techniques,” in Proc. International Mobile, Intelligent, and Ubiquitous Computing Conference (MIUCC), 2024, doi: [https://doi.org/10.1109/MIUCC62295.2024.10783524](https://doi.org/10.1109/MIUCC62295.2024.10783524?utm_source=chatgpt.com). DOI: https://doi.org/10.1109/MIUCC62295.2024.10783524
[20] O. D. Luaces, J. R. Quevedo, J. Díez, C. J. Bahamonde, and A. Alonso, “Binary relevance efficacy for multilabel classification,” Progress in Artificial Intelligence, vol. 1, no. 4, pp. 303–313, 2012, doi: [https://doi.org/10.1007/s13748-012-0030-x](https://doi.org/10.1007/s13748-012-0030-x?utm_source=chatgpt.com). DOI: https://doi.org/10.1007/s13748-012-0030-x
[21] D. J. Touw, “Classifier chain networks for multi-label classification,” Expert Systems with Applications, vol. 286, (ND), 2024, doi: [https://doi.org/10.48550/arXiv.2411.02638](https://doi.org/10.48550/arXiv.2411.02638?utm_source=chatgpt.com). DOI: https://doi.org/10.1016/j.eswa.2025.128048
[22] E. A. Cherman, M. C. Monard, and J. Metz, “Multi-label problem transformation methods: A case study,” CLEI Electronic Journal, vol. 14, no. 1, 2011. (SD) DOI: https://doi.org/10.19153/cleiej.14.1.4
[23] A. G. C. de Sá, P. G. P. Gomes, L. P. (SD), A. A. (SD), and F. (SD), “A robust experimental evaluation of automated multi-label classification methods,” in Proc. Genetic and Evolutionary Computation Conference (GECCO), 2020, pp. 175–183, doi: [https://doi.org/10.1145/3377930.3390231](https://doi.org/10.1145/3377930.3390231?utm_source=chatgpt.com). DOI: https://doi.org/10.1145/3377930.3390231
[24] R. Wirth and J. Hipp, “CRISP-DM: Towards a standard process model for data mining,” in Proc. 4th International Conference on Practical Applications of Knowledge Discovery and Data Mining, 2000, pp. 29–39.
[25] E. Emmanuel et al., “A survey on missing data in machine learning,” Journal of Big Data, vol. 8, no. 1, Art. no. 140, 2021, doi: [https://doi.org/10.1186/s40537-021-00516-9](https://doi.org/10.1186/s40537-021-00516-9?utm_source=chatgpt.com). DOI: https://doi.org/10.1186/s40537-021-00516-9
[26] S. Katoch, S. Chauhan, and V. Kumar, “A review on genetic algorithm: Past, present, and future,” Multimedia Tools and Applications, vol. 80, no. 5, pp. 8091–8126, 2021, doi: [https://doi.org/10.1007/s11042-020-10139-6](https://doi.org/10.1007/s11042-020-10139-6?utm_source=chatgpt.com). DOI: https://doi.org/10.1007/s11042-020-10139-6
[27] T. Akiba, S. Sano, T. Yanase, T. Ohta, and M. Koyama, “Optuna: A next-generation hyperparameter optimization framework,” in Proc. 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), 2019, pp. 2623–2631, doi: [https://doi.org/10.1145/3292500.3330701](https://doi.org/10.1145/3292500.3330701?utm_source=chatgpt.com). DOI: https://doi.org/10.1145/3292500.3330701
[28] Y. Zhang, J. Peng, Y. Sun, and Y. (ND), “Research and application of XGBoost in imbalanced data,” International Journal of Distributed Sensor Networks, vol. 18, no. 6, 2022, doi: [https://doi.org/10.1177/155013292211069](https://doi.org/10.1177/155013292211069?utm_source=chatgpt.com). DOI: https://doi.org/10.1177/15501329221106935
[29] B. Taha Jijo and A. Mohsin Abdulazeez, “Classification based on decision tree algorithm for machine learning,” Journal of Applied Science and Technology Trends, vol. 2, no. 1, pp. 20–28, 2021, doi: [https://doi.org/10.38094/jastt20165](https://doi.org/10.38094/jastt20165?utm_source=chatgpt.com). DOI: https://doi.org/10.38094/jastt20165
[30] A. Parmar, R. Katariya, and V. Patel, “A review on random forest: An ensemble classifier,” in Proc. International Conference on Intelligent Data Communication Technologies and Internet of Things (ICICI), 2018, pp. 758–763, doi: [https://doi.org/10.1007/978-3-030-03146-6_86](https://doi.org/10.1007/978-3-030-03146-6_86?utm_source=chatgpt.com). DOI: https://doi.org/10.1007/978-3-030-03146-6_86
[31] A. Choudhury, A. Mondal, and S. Sarkar, “Searches for the BSM scenarios at the LHC using decision tree-based machine learning algorithms: A comparative study and review of Random Forest, AdaBoost, XGBoost, and LightGBM frameworks,” The European Physical Journal Special Topics, vol. 233, no. 15, pp. 2425–2463, 2024, doi: [https://doi.org/10.1140/epjs/s11734-024-01308-x](https://doi.org/10.1140/epjs/s11734-024-01308-x?utm_source=chatgpt.com). DOI: https://doi.org/10.1140/epjs/s11734-024-01308-x
[32] A. Tharwat, “Classification assessment methods,” Applied Computing and Informatics, vol. 17, no. 1, pp. 168–192, 2021, doi: [https://doi.org/10.1016/j.aci.2018.08.003](https://doi.org/10.1016/j.aci.2018.08.003?utm_source=chatgpt.com). DOI: https://doi.org/10.1016/j.aci.2018.08.003
[33] G. Menardi and N. Torelli, “Training and assessing classification rules with imbalanced data,” Data Mining and Knowledge Discovery, vol. 28, no. 1, pp. 92–122, 2014, doi: [https://doi.org/10.1007/s10618-012-0295-5](https://doi.org/10.1007/s10618-012-0295-5?utm_source=chatgpt.com). DOI: https://doi.org/10.1007/s10618-012-0295-5
[34] C. Schröer, F. Kruse, and J. M. Gómez, “A systematic literature review on applying the CRISP-DM process model,” Procedia Computer Science, vol. 181, pp. 526–534, 2021, doi: [https://doi.org/10.1016/j.procs.2021.01.199](https://doi.org/10.1016/j.procs.2021.01.199?utm_source=chatgpt.com). DOI: https://doi.org/10.1016/j.procs.2021.01.199
