Random Fourier Features Extended Kernel Recursive Least p-Power Algorithm

In this paper, we derive a novel random Fourier features extended kernel recursive least $p$-power (RFF-EW-KRLP) algorithm under the assumption of non-Gaussian impulsive noise. The RFF-EW-KRLP algorithm not only significantly improves convergence rate, steady-state EMSE and tracking ability in the context of impulsive interference, but also reduces the computational complexity replacing the calculation of kernel function with kernel approximation. Simulations are conducted to illustrate the performance benefits of RFF-EW-KRLP related to the typical kernel adaptive filtering algorithms based on the second statistic error criterion in the impulsive noise environment.

[1]  Ignacio Santamaría,et al.  A Sliding-Window Kernel RLS Algorithm and Its Application to Nonlinear Channel Identification , 2006, 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings.

[2]  Xiaoming Chen,et al.  The Kernel Conjugate Gradient Algorithms , 2018, IEEE Transactions on Signal Processing.

[3]  Masahiro Yukawa,et al.  Efficient Dictionary-Refining Kernel Adaptive Filter With Fundamental Insights , 2016, IEEE Transactions on Signal Processing.

[4]  Bernhard Schölkopf,et al.  A Generalized Representer Theorem , 2001, COLT/EuroCOLT.

[5]  Badong Chen,et al.  Quantized Kernel Least Mean Square Algorithm , 2012, IEEE Transactions on Neural Networks and Learning Systems.

[6]  Paul Honeine,et al.  Online Prediction of Time Series Data With Kernels , 2009, IEEE Transactions on Signal Processing.

[7]  N. Aronszajn.  Theory of Reproducing Kernels. , 1950 .

[8]  Nanning Zheng,et al.  Kernel Risk-Sensitive Loss: Definition, Properties and Application to Robust Adaptive Filtering , 2016, IEEE Transactions on Signal Processing.

[9]  Weifeng Liu,et al.  Kernel Adaptive Filtering: A Comprehensive Introduction , 2010 .

[10]  Weifeng Liu,et al.  Extended Kernel Recursive Least Squares Algorithm , 2009, IEEE Transactions on Signal Processing.

[11]  Benjamin Recht,et al.  Random Features for Large-Scale Kernel Machines , 2007, NIPS.

[12]  James R. Zeidler,et al.  Adaptive tracking of linear time-variant systems by extended RLS algorithms , 1997, IEEE Trans. Signal Process..

[13]  Nanning Zheng,et al.  Random fourier feature kernel recursive least squares , 2017, 2017 International Joint Conference on Neural Networks (IJCNN).

[14]  Shukai Duan,et al.  A Class of Weighted Quantized Kernel Recursive Least Squares Algorithms , 2017, IEEE Transactions on Circuits and Systems II: Express Briefs.

[15]  Hao Wu,et al.  Random Fourier Features Based Extended Kernel Recursive Least Squares with Application to fMRI Decoding , 2018, 2018 Chinese Automation Congress (CAC).

[16]  Wei Gao,et al.  Kernel Least Mean $p$-Power Algorithm , 2017, IEEE Signal Processing Letters.

[17]  Sergios Theodoridis,et al.  Online Distributed Learning Over Networks in RKH Spaces Using Random Fourier Features , 2017, IEEE Transactions on Signal Processing.

[18]  Y. Engel.  Kernel Recursive Least Squares , 2004 .

[19]  Sergios Theodoridis,et al.  Efficient KLMS and KRLS algorithms: A random fourier feature perspective , 2016, 2016 IEEE Statistical Signal Processing Workshop (SSP).

[20]  Badong Chen,et al.  Quantized Kernel Recursive Least Squares Algorithm , 2013, IEEE Transactions on Neural Networks and Learning Systems.

[21]  Panayiotis G. Georgiou,et al.  Alpha-Stable Modeling of Noise and Robust Time-Delay Estimation in the Presence of Impulsive Noise , 1999, IEEE Trans. Multim..