Modern approaches to tuning neural network hyperparameters
Keywords:
GA, population, crossover, selectionAbstract
This study presents methods for using genetic algorithms in optimizing the hyper parameters of artificial neural networks. This study provides an analysis of the main types of GA, their mathematical models, and areas of practical application. At the same time, in the process of selecting hyper parameters, the effectiveness of Random Search, Grid Search, and genetic algorithms was compared in the MNIST classification task. The results of the experiment show that GAs provide an optimal balance in terms of calculation time and accuracy and show a higher accuracy compared to the random search method and a shorter calculation time compared to the grid search method. In conclusion, genetic algorithms are considered an effective approach for optimizing the hyper parameters of neural networks, particularly in cases where the parameter space is large and complex
References
1. N. Fayzullo, S. Sariyev and Y. Sherzodjon, "Analyzing the Effectiveness of Ensemble Methods
in Solving Multi-Class Classification Problems," 2025 International Russian Smart Industry
Conference (SmartIndustryCon), Sochi, Russian Federation, 2025, pp. 788-793, doi:
10.1109/SmartIndustryCon65166.2025.10986248
2. Axatov A.R., Nurmamatov M.Q., Nazarov F.M. (2022). Mathematical Models of
Coordination of Population Employment in the Labor Market Ra journal of applied
research[J]. India / – Vol. 8, Issue 2. – Pp. 111–119. https://doi.org/10.47191/rajar/v8i2.09
3. John McCall. (2004) Genetic algorithms for modelling and optimisation School of
Computing[J], Robert Gordon University, Aberdeen, Scotland, UK Received 27 February
2004; received in revised form 7 July 2004. doi:10.1016/j.cam.2004.07.034.
4. A.Rashidov, D. Mardonov, & A. Soliev. Diagnosis of Diabetes Mellitus Based on Artificial
Intelligence Algorithms [C]. 2025 International Russian Smart Industry Conference
(SmartIndustryCon) (in press)
5. Akmal Akhatov, Fayzullo Nazarov, Mekhriddin Nurmamatov, Shokhrukh Sariyev. (2024).
Genetic algorithm application technology in multi-parameter optimization problems AIP
Conf [C]. Proc. 3244, 030025.https://doi.org/10.1063/5.0242074
6. M. Nurmamatov, S. Sariyev and B. Eshonkulov, "Application of Evolutionary Algorithms to
Enhance the Efficiency of Neural Networks and Machine Learning Algorithms," 2025
International Russian Smart Industry Conference (SmartIndustryCon), Sochi, Russian
Federation, 2025, pp. 533-537, doi: 10.1109/SmartIndustryCon65166.2025.10986257
7. Nurmamatov, M., Kulmirzayeva, Z. “Development of an Intelligent System for Optimization
of Employment Information Using Genetic Algorithms” AIP Conference Proceedings, 2024,
3147(1), 040006. doi:10.1063/5.0210279
8. Lakhlifa Sadek, Hamad Talibi Alaoui. (2022). Application of MGA and EGA algorithms on
large-scale linear systems of ordinary differential equations Journal of Computational
ScienceJuly[J] 2022. https://doi.org/10.1016/j.jocs.2022.101719
