Random Fourier Features Multi-Kernel LMS Algorithm

Wei Gao; Meiru Song; Jie Chen; Lingling Zhang

doi:10.1109/ICSPCC50002.2020.9259504

Random Fourier Features Multi-Kernel LMS Algorithm

Wei Gao, Meiru Song, Jie Chen, Lingling Zhang

School of Marine Science and Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

4 Scopus citations

Abstract

Multi-kernel based methods have better and more flexible performance due to more freedom degrees and united features than the mono-kernel methods. In this paper, we present the random Fourier multi-kernel least-mean-square (RFF-MKLMS) algorithm, and derive its analytical models in the mean and mean-square error sense to characterize its transient and steady-state stochastic behaviors. The theoretical predictions consistently match with the simulated learning curves during the transient and steady-state phases, which validate the accuracy of theoretical findings.

Original language	English
Title of host publication	ICSPCC 2020 - IEEE International Conference on Signal Processing, Communications and Computing, Proceedings
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781728172019
DOIs	https://doi.org/10.1109/ICSPCC50002.2020.9259504
State	Published - 21 Aug 2020
Event	2020 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2020 - Macau, China Duration: 21 Aug 2020 → 23 Aug 2020

Publication series

Name	ICSPCC 2020 - IEEE International Conference on Signal Processing, Communications and Computing, Proceedings

Conference

Conference	2020 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2020
Country/Territory	China
City	Macau
Period	21/08/20 → 23/08/20

Keywords

convergence performance analysis
kernel approximation
multi-kernel least-mean-square
Random Fourier features

Access to Document

10.1109/ICSPCC50002.2020.9259504

Cite this

Gao, W., Song, M., Chen, J., & Zhang, L. (2020). Random Fourier Features Multi-Kernel LMS Algorithm. In ICSPCC 2020 - IEEE International Conference on Signal Processing, Communications and Computing, Proceedings Article 9259504 (ICSPCC 2020 - IEEE International Conference on Signal Processing, Communications and Computing, Proceedings). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICSPCC50002.2020.9259504

Gao, Wei ; Song, Meiru ; Chen, Jie et al. / Random Fourier Features Multi-Kernel LMS Algorithm. ICSPCC 2020 - IEEE International Conference on Signal Processing, Communications and Computing, Proceedings. Institute of Electrical and Electronics Engineers Inc., 2020. (ICSPCC 2020 - IEEE International Conference on Signal Processing, Communications and Computing, Proceedings).

@inproceedings{9d6581891eb34f16a33e15b43c2d76d4,

title = "Random Fourier Features Multi-Kernel LMS Algorithm",

abstract = "Multi-kernel based methods have better and more flexible performance due to more freedom degrees and united features than the mono-kernel methods. In this paper, we present the random Fourier multi-kernel least-mean-square (RFF-MKLMS) algorithm, and derive its analytical models in the mean and mean-square error sense to characterize its transient and steady-state stochastic behaviors. The theoretical predictions consistently match with the simulated learning curves during the transient and steady-state phases, which validate the accuracy of theoretical findings.",

keywords = "convergence performance analysis, kernel approximation, multi-kernel least-mean-square, Random Fourier features",

author = "Wei Gao and Meiru Song and Jie Chen and Lingling Zhang",

note = "Publisher Copyright: {\textcopyright} 2020 IEEE.; 2020 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2020 ; Conference date: 21-08-2020 Through 23-08-2020",

year = "2020",

month = aug,

day = "21",

doi = "10.1109/ICSPCC50002.2020.9259504",

language = "英语",

series = "ICSPCC 2020 - IEEE International Conference on Signal Processing, Communications and Computing, Proceedings",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "ICSPCC 2020 - IEEE International Conference on Signal Processing, Communications and Computing, Proceedings",

}

Gao, W, Song, M, Chen, J & Zhang, L 2020, Random Fourier Features Multi-Kernel LMS Algorithm. in ICSPCC 2020 - IEEE International Conference on Signal Processing, Communications and Computing, Proceedings., 9259504, ICSPCC 2020 - IEEE International Conference on Signal Processing, Communications and Computing, Proceedings, Institute of Electrical and Electronics Engineers Inc., 2020 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2020, Macau, China, 21/08/20. https://doi.org/10.1109/ICSPCC50002.2020.9259504

Random Fourier Features Multi-Kernel LMS Algorithm. / Gao, Wei; Song, Meiru; Chen, Jie et al.
ICSPCC 2020 - IEEE International Conference on Signal Processing, Communications and Computing, Proceedings. Institute of Electrical and Electronics Engineers Inc., 2020. 9259504 (ICSPCC 2020 - IEEE International Conference on Signal Processing, Communications and Computing, Proceedings).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Random Fourier Features Multi-Kernel LMS Algorithm

AU - Gao, Wei

AU - Song, Meiru

AU - Chen, Jie

AU - Zhang, Lingling

PY - 2020/8/21

Y1 - 2020/8/21

N2 - Multi-kernel based methods have better and more flexible performance due to more freedom degrees and united features than the mono-kernel methods. In this paper, we present the random Fourier multi-kernel least-mean-square (RFF-MKLMS) algorithm, and derive its analytical models in the mean and mean-square error sense to characterize its transient and steady-state stochastic behaviors. The theoretical predictions consistently match with the simulated learning curves during the transient and steady-state phases, which validate the accuracy of theoretical findings.

AB - Multi-kernel based methods have better and more flexible performance due to more freedom degrees and united features than the mono-kernel methods. In this paper, we present the random Fourier multi-kernel least-mean-square (RFF-MKLMS) algorithm, and derive its analytical models in the mean and mean-square error sense to characterize its transient and steady-state stochastic behaviors. The theoretical predictions consistently match with the simulated learning curves during the transient and steady-state phases, which validate the accuracy of theoretical findings.

KW - convergence performance analysis

KW - kernel approximation

KW - multi-kernel least-mean-square

KW - Random Fourier features

UR - http://www.scopus.com/inward/record.url?scp=85097933415&partnerID=8YFLogxK

U2 - 10.1109/ICSPCC50002.2020.9259504

DO - 10.1109/ICSPCC50002.2020.9259504

M3 - 会议稿件

AN - SCOPUS:85097933415

T3 - ICSPCC 2020 - IEEE International Conference on Signal Processing, Communications and Computing, Proceedings

BT - ICSPCC 2020 - IEEE International Conference on Signal Processing, Communications and Computing, Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2020 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2020

Y2 - 21 August 2020 through 23 August 2020

ER -

Gao W, Song M, Chen J, Zhang L. Random Fourier Features Multi-Kernel LMS Algorithm. In ICSPCC 2020 - IEEE International Conference on Signal Processing, Communications and Computing, Proceedings. Institute of Electrical and Electronics Engineers Inc. 2020. 9259504. (ICSPCC 2020 - IEEE International Conference on Signal Processing, Communications and Computing, Proceedings). doi: 10.1109/ICSPCC50002.2020.9259504

Random Fourier Features Multi-Kernel LMS Algorithm

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this