Assessing the noise immunity and generalization of radial basis function networks

被引：23

作者：

Bernier, JL ^{[1
]}

Díaz, AF ^{[1
]}

Fernández, FJ ^{[1
]}

Cañas, A ^{[1
]}

González, J ^{[1
]}

Martín-Smith, P ^{[1
]}

Ortega, J ^{[1
]}

机构：

[1] Univ Granada, Dept Arquitectura & Tecnol Comp, E-18071 Granada, Spain

来源：

NEURAL PROCESSING LETTERS | 2003年 / 18卷 / 01期

关键词：

generalization; Mean Square Error degradation; noise immunity; perturbation models; Radial Basis Function;

D O I：

10.1023/A:1026275522974

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

In previous work we have derived a magnitude termed the 'Mean Squared Sensitivity' (MSS) to predict the performance degradation of a MLP affected by perturbations in different parameters. The present Letter continues the same line of researching, applying a similar methodology to deal with RBF networks and to study the implications when they are affected by input noise. We obtain the corresponding analytical expression for MSS in RBF networks and validate it experimentally, using two different models for perturbations: an additive and a multiplicative model. We discuss the relationship between MSS and the generalization ability. MSS is proposed as a quantitative measurement to evaluate the noise immunity and generalization ability of a RBFN configuration, giving even more generalization to our approach.

引用

下载

页码：35 / 48

页数：14

共 50 条

[1] Assessing the Noise Immunity and Generalization of Radial Basis Function Networks
J. L. Bernier
A. F. Díaz
F. J. Fernández
A. Cañas
J. González
P. Martín-Smith
J. Ortega
Neural Processing Letters, 2003, 18 : 35 - 48
[2] Generalization Performance of Radial Basis Function Networks
Lei, Yunwen
Ding, Lixin
Zhang, Wensheng
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 2015, 26 (03) : 551 - 564
[3] LEARNING AND GENERALIZATION IN RADIAL BASIS FUNCTION NETWORKS
FREEMAN, JAS
SAAD, D
NEURAL COMPUTATION, 1995, 7 (05) : 1000 - 1020
[4] Generalization and catastrophic forgetting in radial basis function networks
Middleton, N
PROCEEDINGS OF THE NINETEENTH ANNUAL CONFERENCE OF THE COGNITIVE SCIENCE SOCIETY, 1997, : 993 - 993
[5] Improving the Generalization Properties of Radial Basis Function Neural Networks
Bishop, Chris
NEURAL COMPUTATION, 1991, 3 (04) : 579 - 588
[6] A training method for improving the generalization performance of radial basis function networks
Luan, SJ
He, C
Xu, LX
Ma, DS
PROCEEDINGS OF THE 3RD WORLD CONGRESS ON INTELLIGENT CONTROL AND AUTOMATION, VOLS 1-5, 2000, : 859 - 863
[7] RECURRENT RADIAL BASIS FUNCTION NETWORKS FOR ADAPTIVE NOISE CANCELLATION
BILLINGS, SA
FUNG, CF
NEURAL NETWORKS, 1995, 8 (02) : 273 - 290
[8] Active noise control using radial basis function networks
Tokhi, MO
Wood, R
CONTROL ENGINEERING PRACTICE, 1997, 5 (09) : 1311 - 1322
[9] Radial basis function networks: Generalization in over-realizable and unrealizable scenarios
Freeman, JAS
Saad, D
NEURAL NETWORKS, 1996, 9 (09) : 1521 - 1529
[10] An efficient learning algorithm for improving generalization performance of radial basis function neural networks
Wang, ZO
Zhu, T
NEURAL NETWORKS, 2000, 13 (4-5) : 545 - 553

← 1 2 3 4 5 →