A sequential learning algorithm based on adaptive particle filtering for RBF networks

A sequential learning algorithm based on adaptive particle filtering for RBF networks
复制标题

基于自适应粒子滤波的RBF网络顺序学习算法

DOI:
10.1007/s00521-014-1551-y
复制
发表时间:
2014-09
影响因子:
6
通讯作者:
Xiaohong Chen
Xiaohong Chen
中科院分区:
计算机科学3区
文献类型:
--
作者:
Yanhui Xi;Hui Peng(*);Xiaohong Chen

文献摘要

参考文献

相似文献

针对神经网络状态空间模型中未知的过程噪声统计量和局部线性化导致的滤波精度低和发散的问题,提出一种径向基函数(RBF)网络的自适应过程噪声协方差粒子滤波算法。该算法利用扩展卡尔曼粒子滤波算法实现RBF网络权值和中心在时间上依次演化,同时通过对过程噪声协方差矩阵最大化证据密度函数得到过程噪声协方差矩阵。所提出方法的性能通过两个函数逼近问题来评估。实验结果表明,该方法比其他知名训练算法获得了更好的预测精度。
To address the problem of low filtering accuracy and divergence caused by unknown process noise statistics and local linearization in neural network state-space model, this paper proposes an adaptive process noise covariance particle filter algorithm for the radial basis function (RBF) networks. Using the algorithm, the evolution of the weights and centers of RBF networks is achieved sequentially in time by use of the extended Kalman particle filter algorithm, and the process noise covariance matrices are also obtained simultaneously by maximizing the evidence density function with respect to the process noise covariance matrices. Performance of the presented approach is evaluated by two function approximation problems. Experimental results show that the proposed approach obtains better prediction accuracy than other well-known training algorithms.
DOI: --
发表时间: 1997-09
期刊: --
影响因子: --
作者:
J. Freitas;M. Niranjan;A. Gee
通讯作者: J. Freitas;M. Niranjan;A. Gee
DOI: 10.1016/j.apm.2007.09.006
发表时间: 2008-12
影响因子: 5
作者:
F. P. Harter;H. Velho
通讯作者: F. P. Harter;H. Velho
DOI: 10.1109/78.127966
发表时间: 1992-04-01
影响因子: 5.4
作者:
IIGUNI, Y;SAKAI, H;TOKUMARU, H
通讯作者: TOKUMARU, H
DOI: 10.1016/s0065-2458(08)60404-0
发表时间: 1993
期刊: Adv. Comput.
影响因子: --
作者:
Franziska Wulf
通讯作者: Franziska Wulf
DOI: --
发表时间: 1998
影响因子: 2.2
作者:
A. Doucet;S. Godsill;C. Andrieu
通讯作者: A. Doucet;S. Godsill;C. Andrieu