CONCEPT OF ADAPTIVE SELECTION OF STRUCTURED LEARNING ALGORITHMS OF BAYESIAN NETWORKS BASED ON THEIR CHARACTERISTICS

Mariia Voronenko

Abstract


Background. Modern intelligent systems require efficient mechanisms for analysis, prediction, and decision-making. Bayesian networks allow for the efficient representation of causal relationships between variables. The quality and reliability of the constructed network directly correlate with the effectiveness of the final intelligent system.

Materials and Methods. Structural training of a Bayesian network involves determining the structure of a directed acyclic graph in which variables are related to each other. The quality of the structure has a decisive impact on the ability of the model to accurately represent conditional probabilities and on the efficiency of the training algorithms and the reliability of the model. The main problem limiting the structural learning of Bayesian networks is the computational complexity of the model. This fundamental complexity means that for multidimensional problems, it is impossible to perform a complete search of all possible structures and find a global optimum. This forces reliance on heuristic search methods and approximation algorithms and creates a constant need to balance the quality of structure determination and computational resources.

Results and Discussion. The formalized concept of adaptive selection of algorithms for structural learning is based on a systematic analysis of algorithm characteristics and data properties, which allows you to choose the most suitable algorithm for a particular case, optimize the trade-off between model quality and computational resources, and increase the generalizability of the model in practical scenarios.

Conclusion. The proposed concept of adaptive selection of algorithms for structural learning is a timely contribution to the field of stochastic dependence modeling. It successfully translates the process of selecting the optimal algorithm from a routine, heuristic, brute force method to a systematic, multivariate analysis. Its full implementation has the potential to significantly increase the reliability, accuracy, and computational efficiency of building Bayesian models in complex analytical domains.

Keywords: adaptive selection concept, Bayesian network, structured learning algorithm, model accuracy.


Full Text:

PDF

References


[1] Kitson, N.K., Constantinou, A.C., Guo, Z. et al. A survey of Bayesian Network structure learning. Artif Intell Rev 56, 8721–8814 (2023). https://doi.org/10.1007/s10462-022-10351-w.

[2] Wolbrecht E., D’Ambrosio B., Paasch R., Kirby D. Monitoring and diagnosis of a multistage manufacturing process using Bayesian networks. Ai Edam. 1998;14:53–67. https://doi.org/10.1017/S0890060400141058.

[3] Nikovski D. Constructing Bayesian networks for medical diagnosis from vague and partially correct statistics. IEEE Trans. Knowl. Data Eng. 2000;12:509–516. https://doi.org/10.1109/69.868904.

[4] Hosseini S., Barker K. A Bayesian network model for resilience-based supplier selection. Int. J. Prod. Econ. 2016; 180: 68–87. https://doi.org/10.1016/j.ijpe.2016.07.007.

[5] Hosseini S., Barker K. Modeling infrastructure resilience using Bayesian networks: A case study of inland waterway ports. Comput. Ind. Eng. 2016; 93: 252–266. https://doi.org/10.1016/j.cie.2016.01.007.

[6] Hosseini S., Barker K. A Bayesian network model for resilience-based supplier selection. Int. J. Prod. Econ. 2016; 180: 68–87. https://doi.org/10.1016/j.ijpe.2016.07.007.

[7] Dünder E., Cengiz M.A., Koç H. Investigation of the impacts of constraint-based algorithms to the quality of Bayesian network structure in hybrid algorithms for medical studies. J. Adv. Sci. Res. 2014; 5: 8–12. https://sciensage.info/index.php/JASR/article/view/175.

[8] Ziegler V. Approximation algorithms for restricted Bayesian network structures. Inf. Process. Lett. 2008;108:60–63. https://doi.org/10.1016/j.ipl.2008.03.015.

[9] Li S., Zhang J., Sun B., Lei J. An incremental structure learning approach for Bayesian network; Proceedings of the 26th Chinese Control and Decision Conference; Changsha, China. 31 May–2 June 2014; pp. 4817–4822. https://doi.org/10.1109/CCDC.2014.6853036

[10] Chickering D.M., Dan G., Heckerman D. Learning Bayesian networks is NP-hard. Networks. 1994; 112: 121–130.

[11] Xu J.-G., Zhao Y., Chen J., Han C. A structure learning algorithm for Bayesian network using prior knowledge. J. Comput. Sci. Technol. 2015;30:713–724. https://doi.org/10.1007/s11390-015-1556-8.

[12] Amirkhani H., Rahmati M., Lucas P., Hommersom A. Exploiting experts’ knowledge for structure learning of Bayesian networks. IEEE Trans. Pattern Anal. Mach. Intell. 2016; 39: 2154–2170. https://doi.org/10.1109/TPAMI.2016.2636828.

[13] Zhang Y., Zhang W., Xie Y. Improved heuristic equivalent search algorithm based on maximal information coefficient for bayesian network structure learning. Neurocomputing. 2013; 117: 186–195. https://doi.org/10.1016/j.neucom.2013.02.015.

[14] Voronenko M., Mikhailova E., Savina N., Lurie I., and Lytvynenko V., “A Bayesian Network Approach to Measure the Impact of Factors Characterizing the State of Educational Institutions on the Gross Domestic Product”, in Proc. 13th International Conference on Advanced Computer Information Technologies, ACIT 2023, pp. 316–320, 2023. https://doi.org/10.1109/ACIT58437.2023.10275641.

[15] Voronenko M., Zhunissova U., Smailova S., Lytvynenko L., Savina N., Mulesa P., and Lytvynenko V., “Using Bayesian Methods in the Task of Modeling the Patients' Pharmacoresistance Development”, IAPGOŚ, 2, рp.77-82, 2022, http://doi.org/10.35784/iapgos.2968.

[16] Romanko O., Voronenko M., Savina N., Zhorova I., Wójcik W. and Lytvynenko V., "The Use of Static Bayesian Networks for Situational Modeling of National Economy Competitiveness”, in Proc. IEEE International Conference on Advanced Trends in Information Theory (ATIT), pp. 501-505, Kyiv, Ukraine, 2019, https://doi.org/10.1109/ATIT49449.2019.9030515.

[17] Voronenko M., Lurie I., Boskin O., Zhunissova U., Baranenko R. and Lytvynenko V., "Using Bayesian Methods for Predicting the Development of Children Autism”, in Proc. IEEE International Conference on Advanced Trends in Information Theory (ATIT), pp. 525-529, Kyiv, Ukraine, 2019, https://doi.org/10.1109/ATIT49449.2019.9030523.

[18] Li H, Guo H. A Hybrid Structure Learning Algorithm for Bayesian Network Using Experts' Knowledge. Entropy (Basel). 2018 Aug 20; 20 (8): 620. https://doi.org/10.3390/e20080620. PMID: 33265709; PMCID: PMC7513154

[19] Scanagatta, M., Corani, G., de Campos, C. P., & Zaffalon, M. (2018). Approximate structure learning for large Bayesian networks. Machine Learning, 107(8-10), 1209-1227. https://doi.org/10.1007/s10994-018-5701-9.

[20] Stefano Beretta, Mauro Castelli, Ivo Goncalves, Roberto Henriques, Daniele Ramazzotti, “Learning the structure of Bayesian Networks: A quantitative assessment of the effect of different algorithmic schemes”, 2018,
https://doi.org/10.48550/arXiv.1704.08676.

[21] Cheng J., Bell D. A. Liu W. Learning belief networks from data: an information theory based approach // Proceedings of the sixth international conference on information and knowledge management (CIKM 1997), Las Vegas (Nevada), November 10-14. 1997. P. 325-331. https://doi.org/10.1145/266714.266920.

[22] Cheng J., Greiner R., Kelly J., Bell D. A., Liu W. Learning Bayesian networks from data: an information-theory based approach. The artificial intelligence journal (AIJ). 2002. Vol. 137. P. 43-90.

[23] Dempster A. P., Laird N. M., Rubin D. B. Maximum likelihood from incomplete data via the EM algorithm. Journal of the royal statistical society: series B (methodological). 1977. Vol. 39, № 1. P. 1-22. https://doi.org/10.1111/j.2517-6161.1977.tb01600.x.

[24] Freedman D., Kisilev P. Fast mean shift by compact density representation // Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2009. P. 1818-1825. https://doi.org/10.1109/CVPR.2009.5206716.

[25] Friedman N., Murphy K., Russell S. Learning The Structure Of Dynamic Probabilistic Networks // Proc. Fourteenth Conference On Uncertainty In Artificial Intelligence. 1998. https://doi.org/10.48550/arXiv.1301.7374.

[26] Lytvynenko V., Savina N., Krejci J., Voronenko M., Yakobchuk M., and Kryvoruchko O. Bayesian Networks' Development Based on Noisy-MAX Nodes for Modeling Investment Processes in Transport. In Proc. 2-nd International Workshop «Modern Machine Learning Technologies and Data Science» (MoMLeT&DS-2019), pp. 1-10, Shatsk, Ukraine, 2019, http://ceur-ws.org/Vol-2386/.




DOI: http://dx.doi.org/10.30970/eli.32.4

Refbacks

  • There are currently no refbacks.