The analysis of the behavior of a controlled simulation model including the components of the generation of sample trajectories of the regenerating process and the search algorithm is carried out. The formalization of the behavior is proposed in the form of an irreducible return Markov chain, in which the transition probabilities are determined on the basis of comparing the statistical estimates of the functional in the neighboring values of the controlled parameter. It is shown that the transition probabilities are uniquely determined by the functional, which is set by the simulation model of the regenerating process of a complex dynamic system. Stationary probabilities of the constructed Markov chain are found. For an unlimited increase in the discretization of the space of the variable parameter, the convergence of the discrete distribution of stationary probabilities to a continuous random variable is proved. The distribution density of the obtained random variable of the stationary state of the search algorithm model is found. Relationships are obtained between the values of the functional under study and the stationary distribution density of the varied parameters. The analysis showed that the maxima of the distribution density and the function under study coincide, which makes it possible to more accurately estimate the values of the function at the points of the varied parameters closer to the extremum point. As a result, a method of searching for extreme parameters of simulation models of regenerating processes is proposed, based on combining algorithms for generating regenerating processes and algorithms for search optimization.
Pages: 65-73
For citation
Stroganov V.Yu., Chernenky V.M. Analysis of the behavior of search engine optimization algorithms on the models of regenerative processes of complex dynamic systems. Dynamics of complex systems. 2020. T. 14. № 4. Р. 65-73.
DOI: 10.18127/j19997493-202004-07 (In Russian).
References
Ganceva E.A., Kaladze V.A., Kaladze G.A. Dinamicheskie modeli nestacionarnyh sluchajnyh processov. Vestnik VGTU. 2006. T. 2. № 5. S. 4-8. (In Russian).
Egorchev M.V., Tjumencev Ju.V. Adaptivnoe nejrosetevoe modelirovanie dinamicheskih sistem. Sovremennye informacionnye tehnologii i
IT-obrazovanie. 2016. T. 12. № 3-1. S. 195-201 (In Russian).
Egoshin A.B. Postanovka zadachi prognozirovanija vremennogo rjada porozhdaemogo dinamicheskoj sistemoj. Joshkar-Ola: Marijskij gos. tehn. un-t. 2007. S. 136-140. (In Russian).
Kogdenko V.G. Strategicheskoe modelirovanie pribyli kompanii metodom Monte-Karlo. Economic Analysis. 2018. № 17(9). S. 1622-1641
(In Russian).
Kondrat'eva T.N. Prognozirovanie tendencii finansovyh vremennyh rjadov s pomoshh'ju nejronnoj seti LSTM [Jelektronnyj resurs]. Vestnik evrazijskoj nauki. 2017. T. 9. № 4(41). (In Russian).
Krejn M., Lemuan O. Vvedenie v regenerativnyj metod analiza modelej. M.: Nauka. 1982. 104 s. (In Russian).
Kuznecov H.B. Konstruktivnyj metod postroenija «iskusstvennyh» momentov regeneracii. Trudy seminara VNIISI «Teorija slozhnyh sistem i metody ih modelirovanija». M. 1903. S. 15-22. (In Russian).
Kuleshov E.L., Babijchuk I.A. Linejnoe prognozirovanie stacionarnyh sluchajnyh processov pri izvestnom i neizvestnom trende. Avtometrija. 2005. T. 41. № 2. S. 23-35. (In Russian).
Lapshina S.N., Berg D.B., Bazhenov I.A. i dr. Imitacionnye modeli v jekonomike dlja izuchenija scenariev razvitija jekonomicheskih sistem. Jekonomika i upravlenie v mashinostroenii. 2016. № 1. S. 53-55. (In Russian).
Mjet'juz Dzh.G., Fink K.D. Chislennye metody: Ispol'zovanie MATLAB: Per. s angl. L.F. Kozachenko. Pod red. Ju.V. Kozachenko. Izd-e 3. M.: Vil'jams. 2001. 713 s. (In Russian).
Palej A.G., Pollak G.A. Imitacionnoe modelirovanie. Razrabotka imitacionnyh modelej sredstvami iWebsim i Any-Logic: Ucheb. posobie. Sankt-Peterburg: Lan'. 2019. 204 s. (In Russian).
Cyplakov A.A. Vvedenie v prognozirovanie v klassicheskih modeljah vremennyh rjadov. Kvantil'. 2006. № 1. S. 3-19. (In Russian).
Howson C. Successful Business Intelligence, Second Edition: Unlock the Value of BI & Big Data. McGraw-Hill. 2013. 336 p.
Manshour P. Nonlinear correlations in multifractals: Visibility graphs of magnitude and sign series. Chaos: An Interdisciplinary Journal of Non-linear Science. 2020. V. 30(1). P. 013151.
Perslev M., Jensen M.H., Darkner S., et al. U-time: A fully convolutional network for time series segmentation applied to sleep staging.
Advances in Neural Information Processing Systems. 2019. P. 4415-4426.
