Online EM for the Normalized Gaussian Network with Weight-Time-Dependent Updates

被引：0

作者：

Backhus, Jana ^{[1
]}

Takigawa, Ichigaku ^{[1
,2
]}

Imai, Hideyuki ^{[1
]}

Kudo, Mineichi ^{[1
]}

Sugimoto, Masanori ^{[1
]}

机构：

[1] Hokkaido Univ, Grad Sch Informat Sci & Technol, Dept Comp Sci & Informat Technol, Kita Ku, Kita 14 Nishi 9, Sapporo, Hokkaido 0600814, Japan

[2] JST PRESTO, 4-1-8 Honcho, Kawaguchi, Saitama 3320012, Japan

来源：

NEURAL INFORMATION PROCESSING, ICONIP 2016, PT IV | 2016年 / 9950卷

关键词：

Normalized Gaussian networks; Online EM; Local model; Weight-dependent forgetting;

D O I：

10.1007/978-3-319-46681-1_64

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 propose a weight-time-dependent (WTD) update approach for an online EM algorithm applied to the Normalized Gaussian network (NGnet). WTD aims to improve a recently proposed weight-dependent (WD) update approach by Celaya and Agostini. First, we discuss the derivation of WD from an older time-dependent (TD) update approach. Then, we consider additional aspects to improve WD, and by including them we derive the new WTD approach from TD. The difference between WD and WTD is discussed, and some experiments are conducted to demonstrate the effectiveness of the proposed approach. WTD succeeds in improving the learning performance for a function approximation task with balanced and dynamic data distributions.

引用

页码：538 / 546

页数：9

共 6 条

[1] [Anonymous], 1997, Appl. Math.
[2] [Anonymous], P NEUR INF PROC SYST
[3] Online EM with Weight-Based Forgetting
Celaya, Enric
Agostini, Alejandro
[J]. NEURAL COMPUTATION, 2015, 27 (05) : 1142 - 1157
[4] Fast Learning in Networks of Locally-Tuned Processing Units
Moody, John
Darken, Christian J.
[J]. NEURAL COMPUTATION, 1989, 1 (02) : 281 - 294
[5] On-line EM algorithm for the normalized gaussian network
Sato, M
Ishii, S
[J]. NEURAL COMPUTATION, 2000, 12 (02) : 407 - 432
[6] Sato M., 2000, P 7 INT C NEUR INF P, P476

← 1 →