Abstract
The random Fourier feature kernel least mean square (RFF-KLMS) algorithm provides a finite dimensional approximation to the kernel least mean square algorithm with radially symmetric Gaussian kernel. RFF-KLMS was introduced to curb the continuously growing radial basis function (RBF) network which prohibits online application of KLMS. RFF-KLMS assumes a fixed kernel size and the application of the method in nonlinear online regression can be quite tedious because it is not always obvious which kernel size to choose for a particular problem. In this paper, we incorporate a stochastic gradient approach in RFF-KLMS to update the kernel size. The efficacy of the new approach is demonstrated in the online prediction of time series generated from two different chaotic systems. In both examples, the RFF-KLMS algorithm with adaptive kernel size demonstrates very good tracking ability.
The authors would like to acknowledge the financial support from Universiti Sains Malaysia through Research University Grant (RUI) (acc. No. 1001/PMATHS/8011040).
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Takens, F.: Detecting strange attractors in turbulence. Lecture Notes in Math. 898, 366–381 (1981)
Abarbanel, H.D.I.: Analysis of Observed Chaotic Data. Springer-Verlag, New York, Institute for Nonlinear Science (1996)
Huke, J.P., Muldoon, M.R.: Embedding and time series analysis. Math. Today 51(3), 120–123 (2015)
Zhang, S., Han, M., Xu, M.: Chaotic time series online prediction based on improved kernel adaptive filter. In: 2018 International Joint Conference on Neural Networks (IJCNN), pp. 1–6. IEEE, Rio de Janeiro (2018)
Han, M., Zhang, S., Xu, M., Qiu, T., Wang, N.: Multivariate chaotic time series online prediction based on improved kernel recursive least Squares Algorithm. IEEE Trans. Cybern. 49(4), 1160–1172 (2019)
Lu, L., Zhao, H., Chen, B.: Time series prediction using kernel adaptive filter with least mean absolute third loss function. Nonlinear Dyn. 90, 999–1013 (2017)
Garcia-Vega, S., Zeng, X.-J., Keane, J.: Stock rice prediction using kernel adaptive filtering within a stock market interdependence Approach. Available at SSRN (2018). https://doi.org/10.2139/ssrn.3306250
Georga E.I., Principe, J.C., Polyzos, D., Fotiadis, D. I.: Non-linear dynamic modeling of glucose in type 1 diabetes with kernel adaptive filters. In: 2016 38th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), pp. 5897–5900, IEEE, Orlando, FL (2016)
Ouala, S., Nguyen, D., Drumetz, L., Chapron,B., Pascual, A., Collard, F., Gaultier, L., Fablet, R.: Learning latent dynamics for partially-observed chaotic systems. arXiv preprint, arXiv:1907.02452 (2020)
Yin, L., He, Y., Dong, X., Lu, Z.: Adaptive chaotic prediction algorithm of RBF neural network filtering model based on phase Space Reconstruction. J. Comput. 8(6), 1449–1455 (2013)
Feng, T., Yang, S., Han, F.: Chaotic time series prediction using wavelet transform and multi-model hybrid method. J. Vibroeng. 21(7), 1983–1999 (2019)
Kivinen, J., Smola, A., Williamson, R.: Online learning with kernels. In: Advances in Neural Information Processing Systems 14, pp. 785–793. MIT Press (2002)
Comminiello, D., Principe, J.C.: Adaptive Learning Methods for Nonlinear System Modeling. Elsevier (2018)
Chi, M., He, H., Zhang, W.: Nonlinear online classification algorithm with probability. J. Mach. Learn. Res. 20, 33–46 (2011)
Prıncipe, J.C., Liu, W., Haykin, S.: Kernel Adaptive Filtering: A Comprehensive Introduction. 57. John Wiley & Sons (2011)
Paduart, J., Lauwers, L., Swevers, J., Smolders, K., Schoukens, J., Pintelon, R.: Identification of nonlinear systems using Polynomial Nonlinear State Space models. Automatica 46(4), 647–656 (2010)
Liu, W., Pokharel, P.P., Principe, J.C.: The kernel least-mean-square algorithm. IEEE Trans. Signal Process. 56(2), 543–554 (2008)
Engel, Y., Mannor, S., Meir, R.: The kernel recursive least-squares algorithm. IEEE Trans. Signal Process. 52(8), 2275–2285 (2004)
Liu, W., Prıncipe, J.: Kernel affine projection algorithms. EURASIP J. Adv. Signal Process. 2008, 1–12 (2008)
Liu, W., Park, I., Wang, Y., Prıncipe, J.C.: Extended kernel recursive least squares algorithm. IEEE Trans. Signal Process. 57(10), 3801–3814 (2009)
Platt, J.: A resource-allocating network for function interpolation. Neural Comput. 3(2), 213–225 (1991)
Liu, W., Park, I., Prıncipe, J.C.: An information theoretic approach of designing sparse kernel adaptive filters. IEEE Trans. Neural Netw. 20(12), 1950–1961 (2009)
Wang, S., Wang, W., Dang, L., Jiang, Y.: Kernel least mean square based on the Nystrom method. Circuits Syst. Signal Process. 38, 3133–3151 (2019)
Rahimi, A., Recht, B.: Random features for large-scale kernel machines. In: Proceedings of the 21th Annual Conference on Neural Information Processing Systems (ACNIPS), pp. 1177–1184, Vancouver, BC, Canada (2007)
Singh, A., Ahuja, N., Moulin, P.: Online learning with kernels: overcoming the growing sum problem. In: Proceedings of the 2012 IEEE International Workshop on Machine Learning for Signal Process (MLSP), pp. 1–6, Santander, Spain (2012)
Bouboulis, P., Pougkakiotis, S., Theodoridis, S.: Efficient KLMS and KRLS algorithms: a random fourier feature perspective. In: 2016 IEEE Statistical Signal Processing Workshop (SSP), pp. 1–5, Palma de Mallorca (2016)
Xiong, K., Wang, S.: The online random fourier features conjugate gradient algorithm. IEEE Signal Process. Lett. 26(5), 740–744 (2019)
Racine, J.: An efficient cross-validation algorithm for window width selection for nonparametric kernel regression. Commun. Stat. Simul. Comput. 22(4), 1107–1114 (1993)
Cawley, G.C., Talbot, N.L.: Efficient leave-one-out cross-validation of kernel fischer discriminant classifiers. Pattern Recogn. 36(11), 2585–2592 (2003)
An, S., Liu, W., Venkatesh, S.: Fast cross-validation algorithms for least squares support vector machine and kernel ridge regression. Pattern Recogn. 40(8), 2154–2162 (2007)
Hardle, W.: Applied Nonparametric Regression. Volume 5. Cambridge Univ Press (1990)
Herrmann, E.: Local bandwidth choice in kernel regression estimation. J. Comput. Graph. Stat. 6(1), 35–54 (1997)
Chen, B., Liang, J., Zheng, N., PrÃncipe, J.C.: Kernel least mean square with adaptive kernel size. Neurocomput. 191, 95–106 (2016)
Garcia-Vega, S., Zeng, X.-J., Keane, J.: Learning from data streams using kernel least-mean-square with multiple kernel-sizes and adaptive step-size. Neurocomput. 339, 105–115 (2019)
Lorenz, E.N.: Deterministic aperiodic flow. J. Atmos. Sci. 20, 130 (1963)
Zhang, X., Zhu, H., Yao, H.: Analysis of a new three-dimensional chaotic system. Nonlinear Dyn. 67, 335–343 (2012)
Pelikán, E.: Tutorial: forecasting of processes in complex systems for real-world problems. Neural Netw. World 24, 567–589 (2014)
Bolt, E.M.: Regularized forecasting of chaotic dynamical systems. Chaos, Solitons & Fractals. 94, 8–15 (2017)
Kazem, A., Sharifi, E., Hussain, F.K., Saberi, M., Hussain, O.K.: Support vector regression with chaos-based firefly algorithm for stock market price forecasting. Appl. Soft Comput. 13(2), 947–958 (2013)
Brunton, S.L., Proctor, J.L., Kutz, J.N.: Sparse identification of nonlinear dynamics. Proc. National Acad. Sci. 113(15), 3932–3937 (2016)
Scholkopf, B., Smola, A.J.: Learning with Kernels Support Vector Machines, Regularization, Optimization, and Beyond. MIT Press, Cambridge, MA, USA (2001)
Reed, M., Simon, B.: Methods of Modern Mathematical Physics II: Fourier Analysis. Academic Press, Self-Adjointness (1975)
Jones, L.K.: A simple lemma on greedy approximation in Hilbert space and convergence rates for projection pursuit regression and neural network training. Ann. Stat. 20(1), 608–613 (1992)
Liu, Y., Sun, C., Jiang, S.: A kernel least mean square algorithm based on randomized feature networks. Appl. Sci. 8, 458 (2018)
Dong, J., Zheng, Y., Chen, B.: A unified framework of random feature KLMS algorithms and convergence analysis. In: 2018 International Joint Conference on Neural Networks (IJCNN), pp. 1–8, IEEE, Rio de Janeiro (2018)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Ahmad, N.A., Javed, S. (2021). Chaotic Time Series Prediction Using Random Fourier Feature Kernel Least Mean Square Algorithm with Adaptive Kernel Size. In: Mohd, M.H., Misro, M.Y., Ahmad, S., Nguyen Ngoc, D. (eds) Modelling, Simulation and Applications of Complex Systems. CoSMoS 2019. Springer Proceedings in Mathematics & Statistics, vol 359. Springer, Singapore. https://doi.org/10.1007/978-981-16-2629-6_17
Download citation
DOI: https://doi.org/10.1007/978-981-16-2629-6_17
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-16-2628-9
Online ISBN: 978-981-16-2629-6
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)