Advertisement

Analysis and Application of Step Size of RK4 for Performance Measure of Predictability Horizon of Chaotic Time Series

  • Shoya Matsuzaki
  • Kazuya Matsuo
  • Shuichi KurogiEmail author
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11306)

Abstract

So far, we have presented several methods for chaotic time series prediction, and shown performance improvement on predictability horizon. However, we could not have shown the comparison of the performance with other methods. In order to obtain general and absolute performance measure of predictability horizon, this paper analyzes to formulate the relationship between the mean predictability horizon and the step size of the fourth-order Runge-Kutta method, or RK4. By means of using the formula which we have obtained in this article, the step size of RK4 corresponding to the mean predictability horizon achieved by a learning machine can be obtained without executing RK4. We execute numerical experiment of the prediction by several learning machines, and compare the performance by means of the step size of RK4 corresponding to the mean horizon achieved by the learning machines, and we show the effectiveness of the present method.

Keywords

Chaotic time series prediction Analysis and application of step size of RK4 Performance measure of predictability horizon 

References

  1. 1.
    Kurogi, S., Shigematsu, R., Ono, K.: Properties of direct multi-step ahead prediction of chaotic time series and out-of-bag estimate for model selection. In: Loo, C.K., Yap, K.S., Wong, K.W., Teoh, A., Huang, K. (eds.) ICONIP 2014, Part II. LNCS, vol. 8835, pp. 421–428. Springer, Cham (2014).  https://doi.org/10.1007/978-3-319-12640-1_51CrossRefGoogle Scholar
  2. 2.
    Kurogi, S., Toidani, M., Shigematsu, R., Matsuo, K.: Probabilistic prediction of chaotic time series using similarity of attractors and LOOCV predictable horizons for obtaining plausible predictions. In: Arik, S., Huang, T., Lai, W.K., Liu, Q. (eds.) ICONIP 2015. LNCS, vol. 9491, pp. 72–81. Springer, Cham (2015).  https://doi.org/10.1007/978-3-319-26555-1_9CrossRefGoogle Scholar
  3. 3.
    Kurogi, S., Toidani, M., Shigematsu, R., Matsuo, K.: Performance improvement via bagging in probabilistic prediction of chaotic time series using similarity of attractors and LOOCV predictable horizon. Neural Comput. Appl. J. (2017).  https://doi.org/10.1007/s00521-017-3149-7CrossRefGoogle Scholar
  4. 4.
    Kurogi, S., Shimoda, N., Matsuo, K.: Hierarchical clustering of ensemble prediction using LOOCV predictable horizon for chaotic time series. In: Proceedings of IEEE-SSCI 2017, pp. 1789–1795 (2017)Google Scholar
  5. 5.
    Teixeira, J., Reynolds, C.A., Judd, K.K.: Time step sensitivity of nonlinear atmospheric models: numerical convergence, truncation error growth, and ensemble design. J. Atmos. Sci. 64, 175–189 (2007)CrossRefGoogle Scholar
  6. 6.
    Wang, P., Li, J., Li, Q.: Computational uncertainty and the application of a high-performance multiple precision scheme to obtaining the correct reference solution of Lorenz equations. Numer. Algorithms 59(1), 147–159 (2011)MathSciNetCrossRefGoogle Scholar
  7. 7.
    Liao, S., Wang, P.: On the mathematically reliable long-term simulation of chaotic solutions of Lorenz equation in the interval [0, 1000]. Sci. China Phys. Mech. Astron. 57(2), 330–335 (2014)CrossRefGoogle Scholar
  8. 8.
    Aihara, K.: Theories and Applications of Chaotic Time Series Analysis. Sangyo Tosho, Tokyo (2000)Google Scholar
  9. 9.
    Chainer Homepage. https://chainer.org/

Copyright information

© Springer Nature Switzerland AG 2018

Authors and Affiliations

  • Shoya Matsuzaki
    • 1
  • Kazuya Matsuo
    • 1
  • Shuichi Kurogi
    • 1
    Email author
  1. 1.Kyushu Institute of technologyKitakyushu, FukuokaJapan

Personalised recommendations