Configuring RBF neural networks

被引：7

作者：

Sohn, I ^{[1
]}

Ansari, N ^{[1
]}

机构：

[1] New Jersey Inst Technol, Dept Elect & Comp Engn, Newark, NJ 07102 USA

来源：

ELECTRONICS LETTERS | 1998年 / 34卷 / 07期

关键词：

D O I：

10.1049/el:19980469

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

A novel method (based on the characteristics of scatter matrices and frequency-sensitive competitive learning) for training the hidden layer of a radial basis fraction neural network is proposed. The method is demonstrated ro be robust and to outperform the state-of-the-art algorithm.

引用

页码：684 / 685

页数：2

共 5 条

[1] COMPETITIVE LEARNING ALGORITHMS FOR VECTOR QUANTIZATION
AHALT, SC
KRISHNAMURTHY, AK
CHEN, PK
MELTON, DE
[J]. NEURAL NETWORKS, 1990, 3 (03) : 277 - 290
[2] ON THE DISTRIBUTION OF A DEFORMING TRIANGLE
ANSARI, N
DELP, EJ
[J]. PATTERN RECOGNITION, 1990, 23 (12) : 1333 - 1341
[3] Broomhead D. S., 1988, Complex Systems, V2, P321
[4] Fast Learning in Networks of Locally-Tuned Processing Units
Moody, John
Darken, Christian J.
[J]. NEURAL COMPUTATION, 1989, 1 (02) : 281 - 294
[5] RIVAL PENALIZED COMPETITIVE LEARNING FOR CLUSTERING ANALYSIS, RBF NET, AND CURVE DETECTION
XU, L
KRZYZAK, A
OJA, E
[J]. IEEE TRANSACTIONS ON NEURAL NETWORKS, 1993, 4 (04): : 636 - 649

← 1 →