SelfCSHAGA: Self-Configuring Genetic Optimization Algorithm with the Success-History Based Adaptation

Abstract

The paper considers a problem of increasing the genetic optimization algorithm efficiency by using the self-tuning methods that change the algorithm parameters and behavior in finding a solution. It presents an overview of the current and most efficient self-tuning and adaptation methods, their advantages and disadvantages are highlighted. The paper proposes a new algorithm that allows combining best aspects of the separate methods, and is an extended version of the SHAGA method with the improved crossover procedure making it possible to adapt its intensity, apply selective pressure at this stage and use the multi-parent crossover. Within the algorithm framework, the paper proposes various options of the crossover operator and a self-configuration method for the genetic operators based on the SelfCEA approach, which dynamically adjusts probabilities of application depending on their success. The proposed algorithm is tested using the statistical criteria to verify significance of differences in the results and compare with other approaches in the optimization problems with the real and Boolean variables. The testing results of the new genetic algorithm are demonstrating higher efficiency and significant reliability improvement in most test problems

The work was performed with support by the Ministry of Science and Higher Education of the Russian Federation within the framework of the State Task (project no. FSFN-2023-0004)

Please cite this article in English as:

Sherstnev P.A., Semenkin E.S. SelfCSHAGA: self-configuring genetic optimization algorithm with the success-history based adaptation. Herald of the Bauman Moscow State Technical University, Series Instrument Engineering, 2025, no. 2 (151), pp. 122--139 (in Russ.). EDN: TSKBOX

References

[1] Wang Z., Pei Y., Li J. A survey on search strategy of evolutionary multi-objective optimization algorithms. Appl. Sc., 2023, vol. 13, no. 7, art. 4643. DOI: https://doi.org/10.3390/app13074643

[2] Wolpert D.H., Macready W.G. No free lunch theorems for optimization. IEEE Trans. Evol. Comput., 1997, vol. 1, no. 1, pp. 67--82. DOI: https://doi.org/10.1109/4235.585893

[3] Kramer O. Evolutionary self-adaptation: a survey of operators and strategy parameters. Evol. Intel., 2010, vol. 3, no. 2, pp. 51--65. DOI: https://doi.org/10.1007/s12065-010-0035-y

[4] Meyer-Nieberg S., Beyer H.G. Self-adaptation in evolutionary algorithms. In: Parameter setting in evolutionary algorithms. Berlin, Heidelberg, Springer, 2007, pp. 47--75. DOI: https://doi.org/10.1007/978-3-540-69432-8_3

[5] Holland J.H. Genetic algorithms. Sc. Am., 1992, vol. 267, no. 1, pp. 66--72.

[6] Vie A., Kleinnijenhuis A.M., Farmer D.J. Qualities, challenges and future of genetic algorithms: a literature review. arXiv:2011.05277. DOI: https://doi.org/10.48550/arXiv.2011.05277

[7] Alam T., Qamar S., Dixit A., et al. Genetic algorithm: reviews, implementations, and applications. iJEP, 2020, vol. 10, no. 6, pp. 57--76. DOI: https://doi.org/10.36227/techrxiv.12657173

[8] Semenkin E.S., Semenkina M.E. Self-configuring genetic algorithm with modified uniform crossover operator. In: Lecture Notes in Computer Science, Cham, Springer Nature Switzerland, 2012, pp. 414--421. DOI: https://doi.org/10.1007/978-3-642-30976-2

[9] Semenkin E., Semenkina M. Self-configuring genetic programming algorithm with modified uniform crossover. IEEE CEC, 2012. DOI: https://doi.org/10.1109/CEC.2012.6256587

[10] Semenkina M.E. Effectiveness investigation of adaptive evolutionary algorithms for data mining information technology design. Iskusstvennyy intellekt i prinyatie resheniy [Artificial Intelligence and Decision Making], 2013, no. 1, pp. 13--24 (in Russ.). EDN: QBWXYD

[11] Niehaus J., Banzhaf W. Adaption of operator probabilities in genetic programming. In: Genetic Programming. Berlin, Springer-Verlag, 2001, pp. 325--336. DOI: https://doi.org/10.1007/3-540-45355-5_26

[12] Lipinskiy L.V., Kushnareva T.V. Research of self-configurating models and procedures of genetic programming for formation of decision trees in problems of the intelligent data analysis. Vestnik SibGAU im. akademika M.F. Reshetneva [Vestnik SibSAU. Aerospace Tehnologies and Control Systems], 2016, vol. 17, no. 3, pp. 579--586 (in Russ.). EDN: WVPTZH

[13] Sopov A., Karaseva T. The comparison of different PDP-type self-adaptive schemes for the cooperation of GA, DE, and PSO algorithms. ITM Web Conf., 2024, vol. 59, art. 04013. DOI: https://doi.org/10.1051/itmconf/20245904013

[14] Tanabe R., Fukunaga A. Success-history based parameter adaptation for differential evolution. IEEE CEC, 2013, pp. 71--78. DOI: https://doi.org/10.1109/CEC.2013.6557555

[15] Rivera-Lopez R., Mezura-Montes E., Canul-Reich J., et al. An experimental comparison of self-adaptive differential evolution algorithms to induce oblique decision trees. Math. Comput. Appl., 2024, vol. 29, no. 6, p. 103. DOI: https://doi.org/10.3390/mca29060103

[16] Stanovov V., Akhmedova S., Semenkin E. Genetic algorithm with success history based parameter adaptation. IJCCI 2019, 2019, pp. 180--187. DOI: https://doi.org/10.5220/0008071201800187

[17] Yu E.L., Suganthan P.N. Ensemble of niching algorithms. Inform. Sc., 2010, vol. 180, no. 15, pp. 2815--2833. DOI: https://doi.org/10.1016/j.ins.2010.04.008

[18] Mann H.B., Whitney D.R. On a test of whether one of two random variables is stochastically larger than the other. Ann. Math. Statist., 1947, vol. 18, no. 1, pp. 50--60. DOI: https://doi.org/10.1214/aoms/1177730491

[19] Zhu X. Sample size calculation for Mann --- Whitney U test with five methods. Int. J. Clin. Trials, 2021, vol. 8, no. 3, pp. 184--190. DOI: https://doi.org/10.18203/2349-3259.ijct20212840

[20] Stanovov V., Semenkin E. Success rate-based adaptive differential evolution L-SRTDE for CEC 2024 Competition. IEEE CEC, 2024, pp. 1--8. DOI: https://doi.org/10.1109/CEC60901.2024.10611907 ]