The Extended Kernel Adaptive Autoregressive-Moving-Average Algorithm

被引：0

作者：

Dou, Ran ^{[1
]}

Principe, Jose C. ^{[1
]}

机构：

[1] Univ Florida, Dept Elect & Comp Engn, Gainesville, FL 32603 USA

来源：

2022 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN) | 2022年

关键词：

Kernel adaptive filtering; Extended Kalman Filter; time series;

D O I：

10.1109/IJCNN55064.2022.9892233

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

In this paper, we proposed the Extended KAARMA algorithm, which substitutes gradient descent (SGD) by the Extended Kalman (EKF) update equations. Comparing with the stochastic gradient descent method, the EKF method provides a higher rate of convergence. By creating more centers in the memory, it can explore the error space in an efficient manner. Besides, the Extended Kalman method can adjust the learning rate automatically. The decreasing learning rate solves the gradient explosion problem, where a gradient clipping technique is needed in the SGD method instead.

引用

页数：6

共 14 条

[1] Chernodub A.N., 2014, Opt. Mem. Neural Netw, V23, P96, DOI [10.3103/S1060992X14020088, DOI 10.3103/S1060992X14020088]
[2] Hopcroft J.E., 1979, Introduction to Automata Theory, Languages, and Computation
[3] Analysis and improvement of the consistency of extended Kalman filter based SLAM
Huang, Guoquan P.
Mourikis, Anastasios I.
Roumeliotis, Stergios I.
[J]. 2008 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION, VOLS 1-9, 2008, : 473 - 479
[4] Li K., 2015, IEEE T NEURAL NETWOR, V27
[5] Biologically-Inspired Spike-Based Automatic Speech Recognition of Isolated Digits Over a Reproducing Kernel Hilbert Space
Li, Kan
Principe, Jose C.
[J]. FRONTIERS IN NEUROSCIENCE, 2018, 12
[6] Transfer Learning in Adaptive Filters: The Nearest Instance Centroid-Estimation Kernel Least-Mean-Square Algorithm
Li, Kan
Principe, Jose C.
[J]. IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2017, 65 (24) : 6520 - 6535
[7] Lima D., 2017, ENG J AISC, V1, P555
[8] Strictly Positive-Definite Spike Train Kernels for Point-Process Divergences
Park, Il Memming
Seth, Sohan
Rao, Murali
Principe, Jose C.
[J]. NEURAL COMPUTATION, 2012, 24 (08) : 2223 - 2250
[9] Puskorius G. V., 1991, IJCNN-91-Seattle: International Joint Conference on Neural Networks (Cat. No.91CH3049-4), P771, DOI 10.1109/IJCNN.1991.155276
[10] A generalized representer theorem
Schölkopf, B
Herbrich, R
Smola, AJ
[J]. COMPUTATIONAL LEARNING THEORY, PROCEEDINGS, 2001, 2111 : 416 - 426

← 1 2 →