An efficient model selection method for Lotka-Volterra systems based on ABC-PMC and reinforcement learning

i

This article has been updated. A Correction to this article was published on .

i

The Original Article was published on

i

The Original Article was published on

i

This article has been updated

i

This article was retracted on

i

This article has been updated

Abstract

This study provides an efficient, adaptive computational framework for Bayesian model selection when likelihood functions are difficult to handle. To address model selection challenges in scenarios where complex model likelihood functions are difficult to handle, this paper proposes an adaptive approximate Bayesian computation with probability-minimizing computation (RL(Softmax)-ABC-PMC) algorithm that integrates a reinforcement learning (RL) Softmax decision mechanism. This method embeds Softmax agents into the standard ABC-PMC framework, dynamically evaluating the historical performance of candidate models to adaptively focus computational resources on potentially optimal models, thereby significantly enhancing selection efficiency. A numerical experiment is provided to select the best model from three Lotka-Volterra (LV) competition models, including a discretized LV model, a randomized LV model, and a fractional-order LV model, where the algorithm effectively distinguishes subtle differences in their fitting ability and yields reliable parameter estimates. Additionally, we further validate the efficiency of the provided algorithm by selecting the best model from three S-shaped growth curve models. Finally, an empirical analysis using Chinese automotive market sales data further validates the practicality and effectiveness of the proposed method in real-world applications, demonstrating that the RL(Softmax)-ABC-PMC algorithm can automatically identify model structures with stronger explanatory power while maintaining high computational efficiency.

Keywords

model selection / approximate Bayesian computation / LV model / reinforcement learning

CHANGE HISTORY

A Correction to this paper has been published:
https://doi.org

The Retraction Note was updated to reflect reception of statements of disagreement from all authors.

This article has been retracted.Please see the Retraction Notice for more detail:
https://doi.org

References

[1] A. Gelman, J. B. Carlin, H. S. Stern, D. B. Dunson, A. Vehtari, D. B. Rubin, Bayesian Data Analysis , 3^nd edition, Chapman and Hall/CRC, New York, 2013. https://doi.org/10.1201/b16018

[2] M. A. Beaumont, Approximate Bayesian computation in evolution and ecology, Annu. Rev. Ecol. Evol. Syst., 41 (2010), 379-406. https://doi.org/10.1146/annurev-ecolsys-102209-144621

[3] G. Parisi, Nobel lecture: Multiple equilibria, Rev. Mod. Phys., 95 (2023), 030501. https://doi.org/10.1103/RevModPhys.95.030501

[4] K. Csillery, M. Blum, O. E. Gaggiotti, O. Francois, Approximate Bayesian computation (ABC) in practice, Trends Ecol. Evol., 25 (2010), 410-418. https://doi.org/10.1016/j.tree.2010.04.001

[5] J. Marin, P. Pudlo, C. P. Robert, R. J. Ryder, Approximate Bayesian computational methods, Stat. Comput., 22 (2012), 1167-1180. https://doi.org/10.1007/s11222-011-9288-2

[6] P. Marjoram, J. Molitor, V. Plagnol, S. Tavaré, Markov chain Monte Carlo without likelihoods, Proc. Natl. Acad. Sci. U.S.A., 100 (2003), 15324-15328. https://doi.org/10.1073/pnas.0306899100

[7] M. A. Beaumont, W. Zhang, D. J. Balding, Approximate Bayesian computation in population genetics, Genetics, 162 (2002), 2025-2035. https://doi.org/10.1093/genetics/162.4.2025

[8] S. A. Sisson, Y. Fan, M. M. Tanaka, Sequential Monte Carlo without likelihoods, Proc. Natl. Acad. Sci. U.S.A., 104 (2007), 1760-1765. https://doi.org/10.1073/pnas.0607208104

[9] T. Toni, D. Welch, N. Strelkowa, A. Ipsen, M. Stumpf, Approximate Bayesian computation scheme for parameter inference and model selection in dynamical systems, J. R. Soc. Interface, 6 (2009), 187-202. https://doi.org/10.1098/rsif.2008.0172

[10] P. Del Moral, A. Doucet, A. Jasra, An adaptive sequential Monte Carlo method for approximate Bayesian computation, Stat. Comput., 22 (2012), 1009-1020. https://doi.org/10.1007/s11222-011-9271-y

[11] D. T. Frazier, G. M. Martin, C. P. Robert, J. Rousseau, Asymptotic properties of approximate Bayesian computation, Biometrika, 105 (2018), 593-607. https://doi.org/10.1093/biomet/asy027

[12] C. Drovandi, D. T. Frazier, A comparison of likelihood-free methods with and without summary statistics, Stat. Comput., 32 (2022), 42. https://doi.org/10.1007/s11222-022-10092-4

[13] S. Wiqvist, J. Frellsen, U. Picchini, Sequential neural posterior and likelihood approximation, preprint, arXiv:2102.06522.

[14] J. Lueckmann, J. Boelts, D. S. Greenberg, P. J. Goncalves, J. H. Macke, Benchmarking simulation-based inference, preprint, arXiv:2101.04653.

[15] K. Cranmer, J. Brehmer, G. Louppe, The frontier of simulation-based inference, Proc. Natl. Acad. Sci. U.S.A., 117 (2020), 30055-30062. https://doi.org/10.1073/pnas.1912789117

[16] X. Wang, R. P. Kelly, D. J. Warne, C. Drovandi, Preconditioned neural posterior estimation for likelihood-free inference, preprint, arXiv:2404.13557.

[17] C. K. Williams, C. E. Rasmussen, Gaussian Processes for Machine Learning , MIT Press Cambridge, MA, USA, 2006. https://doi.org/10.7551/mitpress/3206.001.0001

[18] W. Zhang, W. Gu, Machine learning for a class of partial differential equations with multi-delays based on numerical Gaussian processes, Appl. Math. Comput., 467 (2024), 128498. https://doi.org/10.1016/j.amc.2023.128498

[19] W. Shi, Study of dynamic systems based on Lotka-Volterra models and machine learning methods, in 2024 5th International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering (ICBAIE), IEEE, (2024), 528-534. https://doi.org/10.1109/ICBAIE63306.2024.11116819

[20] K. N. Dinh, C. Liu, Z. Xiang, Z. Liu, S. Tavare, Approximate Bayesian Computation sequential Monte Carlo via random forests, Stat. Comput., 35 (2025), 219. https://doi.org/10.1007/s11222-025-10748-x

[21] A. B. Abdessalem, N. Dervilis, D. Wagg, K. Worden, Model selection and parameter estimation of dynamical systems using a novel variant of approximate Bayesian computation, Mech. Syst. Signal Process., 122 (2019), 364-386. https://doi.org/10.1016/j.ymssp.2018.12.048

[22] C. P. Robert, J. M. Cornuet, J. M. Marin, N. S. Pillai, Lack of confidence in approximate Bayesian computation model choice, Proc. Natl. Acad. Sci. U.S.A., 108 (2011), 15112-15117. https://doi.org/10.1073/pnas.1102900108

[23] A. Grelaud, J. M. Marin, C. P. Robert, F. Rodolphe, J. F. Taly, ABC methods for model choice in Gibbs random fields, C.R. Math., 347 (2009), 205-210. https://doi.org/10.1016/j.crma.2008.12.009

[24] X. Didelot, R. G. Everitt, A. M. Johansen, D. J. Lawson, ABC Likelihood-free estimation of model evidence, Bayesian Anal., 6 (2011), 49-76. https://doi.org/10.1214/11-BA602

[25] A. Nandy, C. Kumar, D. Mewada, S. Sharma, Bayesian optimization-multi-armed bandit problem, preprint, arXiv:2012.07885.

[26] R. S. Sutton, A. G. Barto, Reinforcement Learning: An Introduction , 2^nd edition, The MIT Press, 2018.

[27] T. G. Ritto, S. Beregi, D. Barton, Reinforcement learning and approximate Bayesian computation for model selection and parameter calibration applied to a nonlinear dynamical system, Mech. Syst. Signal Process., 200 (2023), 110458. https://doi.org/10.1016/j.ymssp.2023.110458

[28] U. Simola, J. Cisewski-Kehe, M. U. Gutmann, J. Corander, Adaptive approximate Bayesian computation tolerance selection, Bayesian Anal., 16 (2021), 397-423. https://doi.org/10.1214/20-BA1211

[29] J. Zhao, M. Yang, Y. Zhao, X. Hu, W. Zhou, H. Li, MCMARL: Parameterizing value function via mixture of categorical distributions for multi-agent reinforcement learning, IEEE Trans. Games, 16 (2024), 556-565. https://doi.org/10.1109/TG.2023.3310150

[30] B. Jin, X. Xu, Machine learning-based forecasts of residential property prices in Hangzhou City, Zhejiang Province, China, Neural Comput. Appl., 37 (2025), 4971-4988. https://doi.org/10.1007/s00521-024-10726-w