Automatic design of convolutional neural network architectures using multi-objective metaheuristics for classification of medical databases
DOI:
https://doi.org/10.61467/2007.1558.2026.v17i4.1213Keywords:
Automatic search of neural architectures, CNN, Evolutionary Computation, Hyperparameters, Metaheuristics, computación evolutiva, hiperparámetros, metaheurísticasAbstract
Convolutional neural networks (CNNs) have been used for different classification tasks, achieving superior performance to machine learning techniques. However, it is necessary to modify the hyperparameters, as their performance depends on the value assigned to them, and the empirical design of CNNs requires a lot of time and effort. Therefore, in this work, we propose a generic methodology, Neural Architecture Search (NAS), to automatically find architectures that classify different medical databases from literature, generating new models that maximize efficiency through the F1 score and minimize the number of parameters. The results of a single-objective (GA) and multi-objective (NSGA-II) genetic algorithm that generated a single architecture for three different medical databases were compared, obtaining an F1 score of 2.9 and 2.831 out of a total of 3 (1 for each database) as the best performance result, respectively. Although results show that the GA got better results than the NSGA-II, the latter took a sixth of the execution time to find the solution compared to the GA. The proposed generic method, with hyperparameter encoding in predefined ranges, allows binary medical databases (e.g., pneumonia or no pneumonia) to be classified efficiently with a single architecture.
Spanish-language metadata / Metadatos en español
Título en español:
Diseño automático de arquitecturas de redes neuronales convolucionales mediante metaheurísticas multiobjetivo para la clasificación de bases de datos médicas
Resumen:
Las redes neuronales convolucionales (CNN) se han utilizado en diferentes tareas de clasificación y han alcanzado un desempeño superior al de las técnicas de aprendizaje automático. Sin embargo, es necesario modificar sus hiperparámetros, ya que su desempeño depende de los valores que se les asignen, y el diseño empírico de las CNN requiere una cantidad considerable de tiempo y esfuerzo. Por lo tanto, en este trabajo se propone una metodología genérica de búsqueda de arquitecturas neuronales —Neural Architecture Search (NAS)— para encontrar automáticamente arquitecturas capaces de clasificar diferentes bases de datos médicas procedentes de la literatura. La metodología genera nuevos modelos que maximizan la eficiencia mediante la puntuación F1 y minimizan el número de parámetros.
Se compararon los resultados de un algoritmo genético de objetivo único (GA) y de un algoritmo genético multiobjetivo (NSGA-II), los cuales generaron una única arquitectura para tres bases de datos médicas diferentes. Como mejores resultados de desempeño, se obtuvieron puntuaciones F1 acumuladas de 2.9 y 2.831, respectivamente, de un máximo total de 3, correspondiente a un valor máximo de 1 para cada base de datos.
Aunque los resultados muestran que el GA obtuvo un mejor desempeño que el NSGA-II, este último requirió una sexta parte del tiempo de ejecución empleado por el GA para encontrar la solución. El método genérico propuesto, basado en la codificación de hiperparámetros dentro de rangos predefinidos, permite clasificar eficientemente bases de datos médicas binarias —por ejemplo, neumonía o ausencia de neumonía— mediante una única arquitectura.
Palabras Claves:
Búsqueda automática de arquitecturas neuronales; CNN; computación evolutiva; hiperparámetros; metaheurísticas.
Smart citations:
https://scite.ai/reports/10.61467/2007.1558.2026.v17i4.1213
Dimensions.
Open Alex.
References
Ait Amou, M., Xia, K., Kamhi, S., & Mouhafid, M. (2022). A novel MRI diagnosis method for brain tumor classification based on CNN and Bayesian optimization. Healthcare, 10(3), Article 494. https://doi.org/10.3390/healthcare10030494
Antczak, K., & Liberadzki, Ł. (2018). Stenosis detection with deep convolutional neural networks. MATEC Web of Conferences, 210, Article 04001. https://doi.org/10.1051/matecconf/201821004001
Atteia, G., Abdel Samee, N., El-Kenawy, E. S. M., & Ibrahim, A. (2022). CNN-hyperparameter optimization for diabetic maculopathy diagnosis in optical coherence tomography and fundus retinography. Mathematics, 10(18), Article 3274. https://doi.org/10.3390/math10183274
Borgli, R. J., Stensland, H. K., Riegler, M. A., & Halvorsen, P. (2019). Automatic hyperparameter optimization for transfer learning on medical image datasets using Bayesian optimization. In 2019 13th International Symposium on Medical Information and Communication Technology (ISMICT) (pp. 1–6). IEEE. https://doi.org/10.1109/ISMICT.2019.8743779
Correa Flórez, C. A., Bolaños Ocampo, R. A., & Molina Cabrera, A. (2008). Algoritmo multiobjetivo NSGA-II aplicado al problema de la mochila [Multiobjective knapsack problem using the NSGA-II algorithm]. Scientia et Technica, 14(39), 206–211.
Deb, K., Pratap, A., Agarwal, S., & Meyarivan, T. (2002). A fast and elitist multiobjective genetic algorithm: NSGA-II. IEEE Transactions on Evolutionary Computation, 6(2), 182–197. https://doi.org/10.1109/4235.996017
Elsken, T., Metzen, J. H., & Hutter, F. (2019). Neural architecture search: A survey. Journal of Machine Learning Research, 20(55), 1–21. https://jmlr.org/papers/v20/18-598.html
Engelbrecht, A. P. (2007). Computational intelligence: An introduction (2nd ed.). Wiley. https://doi.org/10.1002/9780470512517
Franco-Gaona, E., Avila-Garcia, M. S., & Cruz-Aceves, I. (2025). Automatic neural architecture search based on an estimation of distribution algorithm for binary classification of image databases. Mathematics, 13(4), Article 605. https://doi.org/10.3390/math13040605
Gaspar, A., Oliva, D., Cuevas, E., Zaldívar, D., Pérez, M., & Pajares, G. (2021). Hyperparameter optimization in a convolutional neural network using metaheuristic algorithms. In D. Oliva, E. H. Houssein, & S. Hinojosa (Eds.), Metaheuristics in machine learning: Theory and applications (Studies in Computational Intelligence, Vol. 967, pp. 37–59). Springer. https://doi.org/10.1007/978-3-030-70542-8_2
Gil-Rios, M. A., Cruz-Aceves, I., Hernandez-Aguirre, A., Hernandez-Gonzalez, M. A., & Solorio-Meza, S. E. (2024). Improving automatic coronary stenosis classification using a hybrid metaheuristic with diversity control. Diagnostics, 14(21), Article 2372. https://doi.org/10.3390/diagnostics14212372
Goldberg, D. E. (1989). Genetic algorithms in search, optimization, and machine learning. Addison-Wesley.
Hashmi, M. F., Katiyar, S., Keskar, A. G., Bokde, N. D., & Geem, Z. W. (2020). Efficient pneumonia detection in chest X-ray images using deep transfer learning. Diagnostics, 10(6), Article 417. https://doi.org/10.3390/diagnostics10060417
Heinze-Martin, G., Olmedo-Canchola, V. H., Bazán-Miranda, G., Bernard-Fuentes, N. A., & Guízar-Sánchez, D. P. (2018). Los médicos especialistas en México. Gaceta Médica de México, 154(3), 342–351. https://doi.org/10.24875/GMM.18003770
Jang, S., & Son, Y. (2019). Empirical evaluation of activation functions and kernel initializers on deep reinforcement learning. In 2019 International Conference on Information and Communication Technology Convergence (ICTC) (pp. 1140–1142). IEEE. https://doi.org/10.1109/ICTC46691.2019.8939854
Kang, J.-S., Kang, J., Kim, J.-J., Jeon, K.-W., Chung, H.-J., & Park, B.-H. (2023). Neural architecture search survey: A computer vision perspective. Sensors, 23(3), Article 1713. https://doi.org/10.3390/s23031713
Khan, S., Rahmani, H., Shah, S. A. A., & Bennamoun, M. (2018). A guide to convolutional neural networks for computer vision. Morgan & Claypool. https://doi.org/10.2200/S00822ED1V01Y201712COV015
Kilichev, D., & Kim, W. (2023). Hyperparameter optimization for 1D-CNN-based network intrusion detection using GA and PSO. Mathematics, 11(17), Article 3724. https://doi.org/10.3390/math11173724
Krichen, M. (2023). Convolutional neural networks: A survey. Computers, 12(8), Article 151. https://doi.org/10.3390/computers12080151
Kuo, C.-H., Huang, E.-H., Chien, C.-H., & Hsu, C.-C. (2021). FPGA design of enhanced scale-invariant feature transform with finite-area parallel feature matching for stereo vision. Electronics, 10(14), Article 1632. https://doi.org/10.3390/electronics10141632
Lee, S., Kim, J., Kang, H., Kang, D.-Y., & Park, J. (2021). Genetic algorithm based deep learning neural network structure and hyperparameter optimization. Applied Sciences, 11(2), Article 744. https://doi.org/10.3390/app11020744
Li, S., & Ding, W. (2015). Very deep convolutional neural network based image classification using small training sample size. In 2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR) (pp. 730–734). IEEE. https://doi.org/10.1109/ACPR.2015.7486599
Li, Z., Liu, F., Yang, W., Peng, S., & Zhou, J. (2022). A survey of convolutional neural networks: Analysis, applications, and prospects. IEEE Transactions on Neural Networks and Learning Systems, 33(12), 6999–7019. https://doi.org/10.1109/TNNLS.2021.3084827
Melendez, J., van Ginneken, B., Maduskar, P., Philipsen, R. H. H. M., Reither, K., Breuninger, M., Adetifa, I. M. O., Maane, R., Ayles, H., & Sánchez, C. I. (2015). A novel multiple-instance learning-based approach to computer-aided detection of tuberculosis on chest X-rays. IEEE Transactions on Medical Imaging, 34(1), 179–192. https://doi.org/10.1109/TMI.2014.2350539
Monteiro, T. G., Skourup, C., & Zhang, H. (2020). Optimizing CNN hyperparameters for mental fatigue assessment in demanding maritime operations. IEEE Access, 8, 40402–40412. https://doi.org/10.1109/ACCESS.2020.2976601
Mostafa, S. S., Mendonça, F., Ravelo-García, A. G., Juliá-Serdá, G., & Morgado-Dias, F. (2020). Multi-objective hyperparameter optimization of convolutional neural network for obstructive sleep apnea detection. IEEE Access, 8, 129586–129599. https://doi.org/10.1109/ACCESS.2020.3009149
Patel, C. I., Labana, D., Pandya, S., Modi, K., Ghayvat, H., & Awais, M. (2020). Histogram of oriented gradient-based fusion of features for human action recognition in action video sequences. Sensors, 20(24), Article 7299. https://doi.org/10.3390/s20247299
Probst, P., Boulesteix, A.-L., & Bischl, B. (2019). Tunability: Importance of hyperparameters of machine learning algorithms. Journal of Machine Learning Research, 20(53), 1–32. https://jmlr.org/papers/v20/18-444.html
Ren, P., Xiao, Y., Chang, X., Huang, P.-Y., Li, Z., Chen, X., & Wang, X. (2021). A comprehensive survey of neural architecture search: Challenges and solutions. ACM Computing Surveys, 54(4), Article 76. https://doi.org/10.1145/3447582
Simonyan, K., & Zisserman, A. (2014). Very deep convolutional networks for large-scale image recognition [Preprint]. arXiv. https://doi.org/10.48550/arXiv.1409.1556
Simonyan, K., & Zisserman, A. (2015). Very deep convolutional networks for large-scale image recognition. In 3rd International Conference on Learning Representations (ICLR 2015), Conference Track Proceedings. https://arxiv.org/abs/1409.1556
Singh, P., Chaudhury, S., & Panigrahi, B. K. (2021). Hybrid MPSO-CNN: Multi-level particle swarm optimized hyperparameters of convolutional neural network. Swarm and Evolutionary Computation, 63, Article 100863. https://doi.org/10.1016/j.swevo.2021.100863
Stephen, O., Sain, M., Maduh, U. J., & Jeong, D.-U. (2019). An efficient deep learning approach to pneumonia classification in healthcare. Journal of Healthcare Engineering, 2019, Article 4180949. https://doi.org/10.1155/2019/4180949
Talbi, E.-G. (2009). Metaheuristics: From design to implementation. Wiley. https://doi.org/10.1002/9780470496916
Trivedi, M., & Gupta, A. (2022). A lightweight deep learning architecture for the automatic detection of pneumonia using chest X-ray images. Multimedia Tools and Applications, 81(4), 5515–5536. https://doi.org/10.1007/s11042-021-11807-x
Zhang, J., Li, Y., Tai, A., Wen, X., & Jiang, J. (2022). Motion video recognition in speeded-up robust features tracking. Electronics, 11(18), Article 2959. https://doi.org/10.3390/electronics11182959
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 International Journal of Combinatorial Optimization Problems and Informatics
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
