THE MECHANISMS FOR DETERMINING FRACTALITY IN TERMS OF LINGUISTIC MODELING
Nedashkivskyi Yevhen, National Technical University of Ukraine “Kiev Polytechnical Institute”, Kiev, Ukraine
Urgency of the research.Fractaltimeseriesare a class of fractal curves widely used in describing and modeling a variety of events. The use of fractals in modeling time series, including such characteristics of a time series as fractal dimension, allows determining the point at which the system becomes unstable and is ready to move to a new state.
Target setting.Thetheoryoffractals, whenanalyzing market dynamics, allowstaking into accountsuchmarket property as self-organization, so the use of the aforementioned theory to solve the issueof determiningtimeseriesfractalityistopical.
Actual scientific researches and issues analysis. Intherecentstudiesof linguistic modeling modern scholars used the principle of minimum description length within image recognition. A new algorithm of calculating fractal dimension as for the development of the local fractal analysis method that allows satisfactorily solve the problem of identifying local state of the time series under analysis.
Uninvestigated parts of general matters defining. Despitethescaleof scientific research, to date there is no universal mechanism for modeling dynamic processes on thebasisoffractalityoftheobtained time series. Search of this mechanism is one of the most urgent tasks of science today.
The research objective.Theresearchpaper’sobjective is a description and justification of the mechanismfordeterminingfractalityintermsof linguistic modeling. Expand the process of building a linguistic model based on time series of a dynamic process and develop an algorithmfordeterminingfractalityintermsof linguistic modeling.
The statement of basic materials. Theuseoffractals in modeling time series, notably such characteristics of a time series as a fractal dimension, allows determining the point at which the system becomes unstable and is ready to move to a new state. This process unfolds through building a linguistic model based on a time series of the dynamic process. The suggestedalgorithmfordeterminingfractalityintermsof linguistic modeling is implemented in the system unity with the process of modeling long-term memory and results in getting the forecast, including validation (result error evaluation). The mechanism of its implementation consists of six stages. It was found out that the building is determined by linguistic time series data of a cellular automation’s genetic memory. On the basis of the conducted research it was noted that forecasting should be precededby the stage of time series fractal analysis and obtaining additional forecasting information as part of linguistic modeling.
Conclusions. A new approach ofdeterminingthefractalityintermsof linguistic modeling is considered. It is implemented through the mechanism of realization of thealgorithmtodeterminefractalityanditdivided into phases. It is stated that modeling establishes a particular relationship between the model and its original or attributes the model’s properties to the modeled object in the process of special theoreticalanalysisorexperiment.
fractal, linguistic modeling, time series, unstable condition, prognosis, dynamic process
1. Losev, A.F., Vasilenko, I.A. (ed.) (2010). Vvedenie v obshсhuiu teoriiu yazykovykh modelei [Introduction to a General Theory of Linguistic Models]. Moscow: Editorial URSS (in Russian).
2. Shyrokov, V.A., Buhakov, O.V., Hriaznukhina, T.O., Kostyshyn, O.M., Kryhin, M.Yu. (2005). Korpusna linhvistyka[Corpus Linguistics]. Kyiv: Dovira (in Ukrainian).
3. Fu, K.S. (1977). Strukturnye metody v raspoznavanii obrazov [Structural Methods in Recognizing Images] (Trans.). Moscow: Mir (in Russian).
4. Khomskii, N., Miller, Dzh. (2003). Vvedenie v formalnyi analiz estestvennykh yazykov [Introduction to Formal Analysis of Natural Languages]. Moscow: Editorial (in Russian).
5. Potapov, A.S. (2007). Raspoznavanie obrazov i mashinnoe vospriiatie: obshchii pokhod na osnove printsipa minimalnoi dliny opisaniia [Image Recognition and Machine Perception: General Approach on the Basis of Minimum Description Length]. Saint-Petersburg: Politehnika (in Russian).
6. Senkevich, Yu.I. (2007). Algoritmy lingvisticheskogo analiza dinamicheskikh sistem [Algorithms for the Linguistic Analysis of Dynamical Systems]. Voprosy radioelektroniki. Seriia: Elektronno-vychislitelnaia tekhnika– Problems of Radioelectronics. Series: Electronic Computing Machinery, vol. 4, pp. 88–96 (in Russian).
7. Sokolov, A.Yu. (1998). Sintez optimalnogo upravleniia lingvisticheskimi dinamicheskimi sistemami [Synthesis of Optimal Control of Linguistic Dynamic Systems]. Pratsi piatoi ukrainskoi konferentsii z avtomatychnoho upravlinnia (Avtomatyka 98) – Works of the Fifth Ukrainian Conference on Automatic Control, part 1, pp. 356–363 (in Russian).
8. Starchenko, N.V. (2005). Indeks fraktalnosti i lokalnyi analiz khaoticheskikh vremennykh riadov [Fractality Index and Local Analysis of Chaotic Time Series]. Candidate’s thesis. Moscow (in Russian).
9. Baklan, I.V. (2011). Linhvistychne modeliuvannia: osnovy, metody, deiaki prykladni aspekty [Linguistic Modeling: Fundamentals, Techniques, Some Practical Aspects]. Systemni tekhnolohii – System Technology, no. 3, pp. 10–19 (in Ukrainian).
10. Baklan, I.V. (2015). Fraktalni vlastyvosti pry linhvistychnomu modeliuvanni dynamichnykh protsesiv [Fractal Properties in Linguistic Modeling of Dynamic Processes]. Visnyk Khersonskoho natsionalnoho tekhn. un-tu– Bulletin of Kherson National Technical University, vol. 3 (54), pp. 210–213 (in Ukrainian).
11. Baklan, I.V. (2013). Intervalnyi pidkhid do pobudovy linhvistychnoi modeli [Interval Approach to Building a Linguistic Model]. Systemni tekhnolohii – System Technology, vol. 3 (86), pp. 3–8 (in Ukrainian).
12. Sizov, A. A. (2014). Modeli, sposoby i programmnye sredstva podderzhki priniatiia reshenii na osnove prognozirovaniia vremennykh riadov s peremennoi strukturoi [Models, Methods and Software for Decision-Making Support Based on Prediction of Time Series with Variable Structure]. Candidate’s thesis. Moscow (in Russian).
13. Bredihin, A.A., Loskutov, A.Yu. (1998). Vremennye riady s peremennoi dispersiei i finansovye rynki Rossii [Time Series with Variable Variance and Financial Markets of Russia]. Analiz riska – Risk Analysis, issue 1, no. 1, pp. 28–45 (in Russian).
14. Malinetskii, G.G., Kurdiumov, S.P. (2001). Nelineinaia dinamika i problemy prognoza [Nonlinear Dynamics and Forecast Problems]. Vestnik RAN – Bulletin of the Russian Academy of Sciences, issue 71, no. 3, pp. 44–46 (in Russian).
15. Bera, A.K., Higgins, M.L. (1993). ARCH models: properties, astimation and testing. J. Econ. Surveys, vol. 7.
16. Baklan, I.V., Nedashkivskyi, Ye.A. (2016). Linhvistychne modeliuvannia chasovykh riadiv [Linguistic Modeling of Time Series]. Proceedings fromMaterialy IV Vseukrainskoi naukovo-praktychnoi konferentsii studentiv, aspirantiv ta molodykh vchenykh z avtomatychnoho upravlinnia prysviachenoi dniu kosmonavtyky – Materials IV Ukrainian scientific conference of students and young scientists with automatic control of space dedicated to the Day (Kherson, April 12, 2016) (Marasanov V.V. (ed.) et al.). Kherson: KhNTU, pp. 11–14 (in Ukrainian).
17. Koptilov, V.V., Nikitina, F.O. (1966). Chyslo i slovo [The Number and the Word]. Kyiv: Naukova dumka (in Ukrainian).