Multi-Step-Ahead Prediction of Chaotic Time Series: Self-Healing Algorithm for Restoring Values at Non-Predictable Points

Vasilii A. Gromov; Korney K. Tomashchuk; Alexey Rukavishnikov A.

doi:10.48161/qaj.v4n3a912

Authors

Vasilii A. Gromov International Laboratory for Intelligent Systems and Structural Analysis, Faculty of Computer Science, HSE University, Moscow, Russia;
Korney K. Tomashchuk International Laboratory for Intelligent Systems and Structural Analysis, Faculty of Computer Science, HSE University, Moscow, Russia;
Alexey A. Rukavishnikov School of Data Analysis and Artificial Intelligence, Faculty of Computer Science, HSE University, Moscow, Russia.

DOI:

https://doi.org/10.48161/qaj.v4n3a912

Abstract

Problem: The study proposes a new algorithm for multi-step-ahead prediction of chaotic time series within the framework of the clustering-based forecasting paradigm. The introduction of the concept of non-predictable points enabled to avoid the exponential growth of the prediction error as a function of the number of steps ahead, which made it possible to develop algorithms that predict many Lyapunov times (and many steps) ahead – the price for this turned out to be that some points remained non-predictable. In this work, we say that it is necessary to be of great help regarding the exemption and development of the scientific university. Method: This study proposes a self-healing algorithm, which is an iterative algorithm that takes as input the forecasts produced by the underlying prediction algorithm. At each iteration, the self-healing algorithm finds new possible predicted values, updates the status of the points from predictable to non-predictable or vice versa, and calculates new single predicted values for the predictable points. The study proposes several new algorithms for calculating a single prediction value and algorithms for determining non-predictable points. We have studied new publications of parameter estimates, importance monitoring, and predictive features with more views and size objectives. Results: In the results of our work, the new tools in specific indicators: RMSE was increased from 0.11 to 0.06, and MAPE was reduced from 0.38 to 0.04. Conclusion: Research has been conducted on the selection of parameters for the self-healing algorithm, its assessment, and the prediction quality compared to the existing prediction algorithm.

Downloads

Download data is not yet available.

References

Gromov, V. A., & Baranov, P. S. (2023). Prediction after a horizon of predictability: Nonpredictable points and partial multistep prediction for chaotic time series. Complex, 1–21.

Malinetsky, G. G., & Potapov, A. B. (2000). Modern problems of nonlinear dynamics. Editorial URSS: Moscow, Russia.

Meng, X., & Yang, T. (2021). Entanglement-structured LSTM boosts chaotic time series forecasting. Entropy, 23, 1491.

Bao, Y., Xiong, T., & Hu, Z. (2014). Multi-step-ahead time series prediction using multiple-output support vector regression. Neurocomputing, 129, 482–493.

Assaad, M., Boné, R., & Cardot, H. (2006). Predicting chaotic time series by boosted recurrent neural networks. In Lecture Notes in Computer Science (pp. 831–840). Springer: Berlin, Germany.

Taieb, S. B., Bontempi, G., Atiya, A. F., & Sorjamaa, A. (2012). A review and comparison of strategies for multi-step ahead time series forecasting based on the NN5 forecasting competition. Expert Systems with Applications, 39, 7067–7083.

Sorjamaa, A., Hao, J., Reyhani, N., Ji, Y., & Lendasse, A. (2007). Methodology for long-term prediction of time series. Neurocomputing, 70, 2861–2869.

Bontempi, G. (2008). Long term time series prediction with multi-input multi-output local learning. Proceedings of the 2nd European Symposium on Time Series Prediction (TSP), 145–154.

Taieb, S. B., Bontempi, G., Sorjamaa, A., & Lendasse, A. (2009). Long-term prediction of time series by combining direct and MIMO strategies. Proceedings of the 2009 International Joint Conference on Neural Networks, 3054-3061.

Chandra, R., Goyal, S., & Gupta, R. (2021). Evaluation of deep learning models for multi-step ahead time series prediction. IEEE Access, 9, 83105–83123.

Gromov, V. A., & Dang, Q. N. (2023). Semantic and sentiment trajectories of literary masterpieces.

Gromov, V. A., & Konev, A. S. (2016). Precocious identification of popular topics on Twitter with the employment of predictive clustering. Neural Computing and Applications, 28, 3317–3322.

Gromov, V. (2021). Chaotic time series prediction: Run for the horizon. In Tools and methods of program analysis (pp. 29-43). Springer: Cham, Switzerland.

Gromov, V. A., & Shulga, A. N. (2012). Chaotic time series prediction with employment of ant colony optimization. Expert Systems with Applications, 39, 8474–8478.

Takens, F. (1981). Detecting strange attractors in turbulence. In Lecture Notes in Mathematics (pp. 366–381). Springer: Berlin, Germany.

Gromov, V. A., & Borisenko, E. A. (2015). Chaotic time series prediction and clustering methods. Neural Computing and Applications, 2, 307–315.

Ester, M., Kriegel, H.-P., Sander, J., & Xu, X. (1996). A density-based algorithm for discovering clusters in large spatial databases with noise. Knowledge Discovery and Data Mining, 226–231.

Wishart, D. (1969). Numerical classification method for deriving natural classes. Nature, 221, 97–98.

Lapko, A. V., & Chentsov, S. V. (2000). Nonparametric information processing systems. Nauka: Novosibirsk, Russia.