Abstract
The kernel recursive least squares (KRLS), a nonlinear counterpart of the famed RLS algorithm, performs linear regression in a high-dimensional feature space induced by a Mercer kernel. Despite the growing interest in the KRLS for nonlinear signal processing, the presence of outliers in the estimation data causes the resulting predictor’s performance to deteriorate considerably. Bearing this in mind, we introduce an approach to amalgamate the kernel-based learning framework that gives rise to the KRLS algorithm with the robust regression framework of M-estimators with the aim of building an outlier-robust variant for the KRLS. Initially, we develop the theoretical aspects of the proposed algorithm and then analyze its behavior in nonlinear system identification problems using synthetic and real-world datasets (including a large-scale one) contaminated with outliers. The obtained results indicate that the robust variant of the KRLS algorithm consistently outperforms the state-of-the-art in robust adaptive filtering algorithms, as the amount of outliers in the data increases.
Similar content being viewed by others
Notes
A signal processing task in which an inverse model of a communication channel (e.g., in cellular systems) is built and its parameters are adapted in an online fashion to compensate for “intersymbol interference”, a phenomenon caused by signal pulse smearing in dispersive media.
A collection of atoms which can represent each training data by a linear combination of the atoms.
References
Bakker, R., Schouten, J.C., Coppens, M.O., Takens, F., Giles, C.L., van den Bleek, M.: Robust learning of chaotic attractors. In: Solla, S.A., Leen, T.K., Müller, K.R. (eds.) Advances in Neural Information Processing Systems, vol. 12, pp. 879–885. MIT Press (2000)
Bhotto, M.Z.A., Antoniou, A.: Robust recursive least-squares adaptive-filtering algorithm for impulsive-noise environments. IEEE Signal Process. Lett. 18(3), 185–188 (2011)
Billings, S.A.: Nonlinear System Identification NARMAX Methods in the Time, Frequency and Spatio-Temporal Domains, 1st edn. Wiley, London (2013)
Billings, S.A., Voon, W.S.F.: Correlation based model validity tests for non-linear models. Int. J. Control 44(1), 235–244 (1986)
Cai, Y., Wang, H., Ye, X., Fan, Q.: A multiple-kernel LSSVR method for separable nonlinear system identification. J. Control Theory Appl. 11(4), 651–655 (2013)
Campbell, C.: Kernel methods: a survey of current techniques. Neurocomputing 48(1–4), 63–84 (2002)
Chan, S.C., Zou, Y.X.: A recursive least M-estimate algorithm for robust adaptive filtering in impulsive noise: fast algorithm and convergence performance analysis. IEEE Trans. Signal Process. 52(4), 975–991 (2004)
Chaudhary, N.I., Raja, M.A.Z.: Identification of Hammerstein nonlinear ARMAX systems using nonlinear adaptive algorithms. Nonlinear Dyn. 79(2), 1385–1397 (2015)
De Brabanter, K., De Brabanter, J., Suykens, J.A., Vandewalle, J., De Moor, B.: Robustness of kernel based regression: influence and weight functions. In: The 2012 International Joint Conference on Neural Networks (IJCNN), pp. 1–8 (2012)
De Brabanter, K., Pelckmans, K., De Brabanter, J., Debruyne, M., Suykens, J.A., Hubert, M., De Moor, B.: Robustness of kernel based regression: a comparison of iterative weighting schemes. In: Artificial Neural Networks–ICANN 2009, pp. 100–110. Springer (2009)
Debruyne, M., Christmann, A., Hubert, M., Suykens, J.A.K.: Robustness and stability of reweighted kernel based regression. Technical Report 06-09, Department of Mathematics, Katholiek University of Leuven, Leuven, Belgium (2008)
Engel, Y., Mannor, S., Meir, R.: Sparse online greedy support vector regression. In: Elomaa, T., Mannila, H., Toivonen, H. (eds.) Proceedings of the 13th European Conference on Machine Learning (ECML’2002), pp. 84–96 (2002)
Engel, Y., Mannor, S., Meir, R.: The kernel recursive least-squares algorithm. IEEE Trans. Signal Process. 52(8), 2275–2285 (2004)
Espinoza, M., Pelckmans, K., Hoegaerts, L., Suykens, J.A., De Moor, B.: A comparative study of LS-SVMs applied to the silverbox identification problem. In: Proceedings of the 6th IFAC Symposium on Nonlinear Control Systems (NOLCOS’2004) (2004)
Falck, T., Dreesen, P., De Brabanter, K., Pelckmans, K., De Moor, B., Suykens, J.A.K.: Least-squares support vector machines for the identification of Wiener–Hammerstein systems. Control Eng. Pract. 20(11), 1165–1174 (2012)
Fan, H., Song, Q.: A sparse kernel algorithm for online time series data prediction. Expert Syst. Appl. 40, 2174–2181 (2013)
Farahmand, S., Giannakis, G.B.: Robust RLS in the presence of correlated noise using outlier sparsity. IEEE Trans. Signal Process. 60(6), 3308–3313 (2012)
Fox, J.: An R and S-Plus Companion to Applied Regression, 1st edn. SAGE (2002)
Golub, G.H., Van Loan, C.F.: Matrix Computations, vol. 3. JHU Press, Baltimore (2012)
Haykin, S.: Adaptive Filter Theory, 5th edn. Prentice Hall, Englewood Cliffs (2013)
Huber, P.J., Ronchetti, E.M.: Robust Statistics. Wiley, London (2009)
Huber, P.J., et al.: Robust estimation of a location parameter. Ann. Math. Stat. 35(1), 73–101 (1964)
Jäkel, F., Schölkopf, B., Wichmann, F.A.: A tutorial on kernel methods for categorization. J. Math. Psychol. 51(6), 343–358 (2007)
Jang, J.S.: ANFIS: adaptive-network-based fuzzy inference system. IEEE Trans. Syst. Man Cybern. 23(3), 665–685 (1993)
Kocijan, J., Girard, A., Banko, B., Murray-Smith, R.: Dynamic systems identification with Gaussian processes. Math. Comput. Model. Dyn. Syst. 11(4), 411–424 (2005)
Li, X., Zhou, L., Sheng, J., Ding, R.: Recursive least squares parameter estimation algorithm for dual-rate sampled-data nonlinear systems. Nonlinear Dyn. 76(2), 1327–1334 (2014)
Liu, B., Wang, Z.: On system identification based on online least squares support vector machine. In: International Conference on Intelligent Systems and Knowledge Engineering. Atlantis Press (2007)
Liu, W., Park, I., Principe, J.: An information theoretic approach of designing sparse kernel adaptive filters. IEEE Trans. Neural Netw. 20(12), 1950–1961 (2009)
Liu, W., Park, I., Wang, Y., Principe, J.: Extended kernel recursive least squares algorithm. IEEE Trans. Signal Process. 57(10), 3801–3804 (2009)
Liu, W., Pokharel, P., Principe, J.: The kernel least mean square algorithm. IEEE Trans. Signal Process. 56(2), 543–554 (2008)
Liu, Y., Chen, J.: Correntropy kernel learning for nonlinear system identification with outliers. Ind. Eng. Chem. Res. 53, 5248–5260 (2014)
Liu, Y., Wang, H., Yu, J., Li, P.: Selective recursive kernel learning for online identification of nonlinear systems with NARX form. J. Process Control 20(2), 181–194 (2010)
Ljung, L.: System Identification: Theory for the User, 2nd edn. Prentice Hall, Englewood Cliffs (1999)
Ljung, L.: Recursive identification algorithms. Circuits Syst. Signal Process. 21(1), 57–68 (2002)
Mattos, C.L.C., Damianou, A., Barreto, G.A., Lawrence, N.D.: Latent autoregressive Gaussian processes models for robust system identification. In: Proceedings of the 11th IFAC Symposium on Dynamics and Control of Process Systems (DYCOPS’2016) (2016, to appear)
Naderahmadian, Y., Tinati, M.A., Beheshti, S.: Generalized adaptive weighted recursive least squares dictionary learning. Sig. Process. 118, 89–96 (2016)
Narendra, K.S., Parthasarathy, K.: Identification and control of dynamical systems using neural networks. IEEE Trans. Neural Netw. 1(1), 4–27 (1990)
Papageorgiou, G., Bouboulis, P., Theodoridis, S.: Robust linear regression analysis—a greedy approach. IEEE Trans. Signal Process. 63(15), 3872–3887 (2015)
Pillonetto, G., Dinuzzob, F., Chen, T., De Nicolao, G., Ljung, L.: Kernel methods in system identification, machine learning and function estimation: a survey. Automatica 50, 657–682 (2014)
Pintelon, R., Schoukens, J.: System Identification: A Frequency Domain Approach. Wiley, London (2012)
Platt, J.: A resource-allocating network for function interpolation. Neural Comput. 3(2), 213–225 (1991)
Principe, J., Rathie, A., Kuo, J.M.: Prediction of chaotic time series with neural networks and the issue of dynamic modeling. Int. J. Bifurc. Chaos 2(4), 989–996 (1992)
Quan, T., Liu, X., Liu, Q.: Weighted least squares support vector machine local region method for nonlinear time series prediction. Appl. Soft Comput. 10(2), 562–566 (2010)
Richard, C., Bermudez, J.C.M., Honeine, P.: Online prediction of time series data with kernels. IEEE Trans. Signal Process. 57(3), 1058–1066 (2009)
Rojo-Alvarez, J.L., Martinez-Ramon, M., Munoz-Maric, J., Camps-Valls, G.: A unified SVM framework for signal estimation. Digit. Signal Proc. 26, 1–20 (2014)
Rousseeum, P.J., Leroy, A.M.: Robust Regression and Outlier Detection, 1st edn. Wiley, London (1987)
Saide, C., Lengellé, R., Honeine, P., Achkar, R.: Online kernel adaptive algorithms with dictionary adaptation for MIMO models. IEEE Signal Process. Lett. 20(5), 535–538 (2013)
Saide, C., Lengelle, R., Honeine, P., Richard, C., Achkar, R.: Nonlinear adaptive filtering using kernel-based algorithms with dictionary adaptation. Int. J. Adapt. Control Signal Process. 29(11), 1391–1410 (2015)
Santos, J.D.A., Mattos, C.L.C., Barreto, G.A.: Performance evaluation of least squares SVR in robust dynamical system identification. In: Proceedings of the 13th International Work-Conference on Artificial Neural Networks (IWANN’2015), pp. 422–435 (2015)
Saunders, C., Gammerman, A., Vovk, V.: Ridge regression learning algorithm in dual variables. In: Proceedings of the 15th International Conference on Machine Learning (ICML’98), pp. 515–521 (1998)
Schölkopf, B., Smola, A.J.: Learning With Kernels: Support Vector Machines, Regularization, Optimization, and Beyond. MIT Press, Cambridge (2002)
Schoukens, J., Németh, J.G., Crama, P., Rolain, Y., Pintelon, R.: Fast approximate identification of nonlinear systems. Automatica 39(7), 1267–1274 (2003)
Sjöberg, J., Zhang, Q., Ljung, L., Benveniste, A., Delyon, B., Glorennec, P.Y., Hjalmarsson, H., Juditsky, A.: Nonlinear black-box modeling in system identification: a unified overview. Automatica 31(12), 1691–1724 (1995)
Smola, A.J., Schölkopf, B.: A tutorial on support vector regression. Stat. Comput. 14(3), 199–222 (2004)
Suykens, J.A.K., De Brabanter, J., Lukas, L., Vandewalle, J.: Weighted least squares support vector machines: robustness and sparse approximation. Neurocomputing 48(1), 85–105 (2002)
Suykens, J.A.K., Van Gestel, T., De Brabanter, J., De Moor, B., Vandewalle, J.: Least Squares Support Vector Machines. World Scientific, Singapore (2002)
Takagi, T., Sugeno, M.: Fuzzy identification of systems and its application to modeling and control. IEEE Trans. Syst. Man Cybern. 15(1), 116–132 (1985)
Tang, H.S., Xue, S.T., Chen, R., Sato, T.: Online weighted LS-SVM for hysteretic structural system identification. Eng. Struct. 28(12), 1728–1735 (2006)
Taouali, O., Elaissi, E., Messaoud, H.: Design and comparative study of online kernel methods identification of nonlinear system in RKHS space. Artif. Intell. Rev. 37(4), 289–300 (2012)
Tenreiro Machado, J.A., Babaei, A., Moghaddam, B.P.: Highly accurate scheme for the Cauchy problem of the generalized Burgers–Huxley equation. Acta Polytech. Hung. 13(6), 183–195 (2016)
Van Vaerenbergh, S., Vía, J., Santamaría, I.: Nonlinear system identification using a new sliding-window kernel RLS algorithm. J. Commun. 2(3), 1–8 (2007)
Van Vaerenbergh, S., Vía, J., Santamaría, I.: Kernel recursive least-squares tracker for time-varying regression. IEEE Trans. Neural Netw. Learn. Syst. 23(8), 1313–1326 (2012)
Vapnik, V.: The Nature of Statistical Learning Theory, 1st edn. Springer, Berlin (1995)
Vapnik, V.: Statistical Learning Theory, 1st edn. Wiley, London (1998)
Wedderburn, R.W.: Quasi-likelihood functions, generalized linear models, and the Gauss–Newton method. Biometrika 61(3), 439–447 (1974)
Widrow, B., Lehr, M.: Noise canceling and channel equalization. In: Arbib, M.A. (ed.) Handbook of Brain Theory and Neural Networks, pp. 648–650. MIT Press, Cambridge (1995)
Widrow, B., Walach, E.: Adaptive Inverse Control: A Signal Processing Approach. Wiley-IEEE Press, London (2007)
Wills, A., Schön, T.B., Ljung, L., Ninness, B.: Identification of Hammerstein–Wiener models. Automatica 49(1), 70–81 (2013)
Wu, Z., Shi, J., Zhang, X., Ma, W., Chen, B.: Kernel recursive maximum correntropy. Signal Process. 117, 11–16 (2015)
Yang, X.J.: A new integral transform operator for solving the heat-diffusion problem. Appl. Math. Lett. 64, 193–197 (2017)
Yang, X.J., Gao, F.: A new technology for solving diffusion and heat equations. Therm. Sci. 21, 133–140 (2017)
Young, P.C.: Recursive Estimation and Time-series Analysis: An Introduction for the Student and Practitioner. Springer, Berlin (2011)
Zhang, B., Billings, S.A.: Identification of continuous-time nonlinear systems: the nonlinear difference equation with moving average noise (NDEMA) framework. Mech. Syst. Signal Process. 60, 810–835 (2015)
Zhang, B., Billings, S.A.: Volterra series truncation and kernel estimation of nonlinear systems in the frequency domain. Mech. Syst. Signal Process. 84–A, 39–57 (2017)
Zhang, J., Pang, Y.: Pipelined robust \(m\)-estimate adaptive second-order volterra filter against impulsive noise. Digit. Signal Process. 26, 71–80 (2014)
Zhang, K., Fei, M., Wu, J., Zhang, P.: Fast prediction model based big data system identification. In: Chinese Automation Congress (CAC), pp. 465–469. IEEE (2013)
Zhang, Y.Q., Wang, X.Y.: Spatiotemporal chaos in mixed linear-nonlinear coupled logistic map lattice. Physica A 402, 104–118 (2014)
Zhang, Y.Q., Wang, X.Y., Liu, L.Y., He, Y., Liu, J.: Spatiotemporal chaos of fractional order logistic equation in nonlinear coupled lattices. Commun. Nonlinear Sci. Numer. Simul. 52, 52–61 (2017)
Zhao, S., Chen, B., Principe, J.C.: Kernel adaptive filtering with maximum correntropy criterion. In: Proceedings of the 2011 International Joint Conference on Networks (IJCNN’2011), pp. 2012–2017 (2011)
Zhu, P.P., Chen, B., Principe, J.C.: A novel extended kernel recursive least squares algorithm. Neural Netw. 32, 349–357 (2012)
Zou, Y., Chan, S., Ng, T.: A recursive least M-estimate (RLM) adaptive filter for robust filtering in impulse noise. IEEE Signal Process. Lett. 7(11), 324–326 (2000)
Zoubir, A.M., Koivunen, V., Chakhchoukh, Y., Muma, M.: Robust estimation in signal processing: a tutorial-style treatment of fundamental concepts. IEEE Signal Process. Mag. 29(4), 61–80 (2012)
Acknowledgements
The authors thank the financial support of Federal Institute of Ceará (IFCE), NUTEC and CNPq (Grant 309451/2015-9).
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Santos, J.D.A., Barreto, G.A. An outlier-robust kernel RLS algorithm for nonlinear system identification. Nonlinear Dyn 90, 1707–1726 (2017). https://doi.org/10.1007/s11071-017-3760-2
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11071-017-3760-2