A new technique for adjusting the learning rate of RPEM algorithm automatically

Xing Ming Zhao; Yiu Ming Cheung; Luonan Chen; Kazuyuki Aihara

A new technique for adjusting the learning rate of RPEM algorithm automatically

Xing Ming Zhao^*, Yiu Ming Cheung, Luonan Chen, Kazuyuki Aihara

^*Corresponding author for this work

Department of Computer Science

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

2 Citations (Scopus)

Abstract

Recently, a new Rival Penalized Expectation Maximization (RPEM) algorithm has been proposed for estimating the parameters of the normal mixture model, meanwhile determining the number of classes automatically. The RPEM is an adaptive algorithm utilizing a small constant learning rate. To speed up its convergence speed, this paper proposes a new method to dynamically adjust the learning rate of the RPEM algorithm on line. The numerical results have shown the promising results of the proposed algorithm.

Original language	English
Title of host publication	Proceedings of the 12th International Symposium on Artificial Life and Robotics, AROB 12th'07
Pages	597-600
Number of pages	4
Publication status	Published - 2007
Event	12th International Symposium on Artificial Life and Robotics, AROB 12th'07 - Oita, Japan Duration: 25 Jan 2007 → 27 Jan 2007

Publication series

Name	Proceedings of the 12th International Symposium on Artificial Life and Robotis, AROB 12th'07

Conference

Conference	12th International Symposium on Artificial Life and Robotics, AROB 12th'07
Country/Territory	Japan
City	Oita
Period	25/01/07 → 27/01/07

Scopus Subject Areas

Artificial Intelligence
Computer Vision and Pattern Recognition
Human-Computer Interaction

Cite this

@inproceedings{9cb15f1028404648bf57c9495e3dc35c,

title = "A new technique for adjusting the learning rate of RPEM algorithm automatically",

abstract = "Recently, a new Rival Penalized Expectation Maximization (RPEM) algorithm has been proposed for estimating the parameters of the normal mixture model, meanwhile determining the number of classes automatically. The RPEM is an adaptive algorithm utilizing a small constant learning rate. To speed up its convergence speed, this paper proposes a new method to dynamically adjust the learning rate of the RPEM algorithm on line. The numerical results have shown the promising results of the proposed algorithm.",

author = "Zhao, {Xing Ming} and Cheung, {Yiu Ming} and Luonan Chen and Kazuyuki Aihara",

year = "2007",

language = "English",

isbn = "9784990288013",

series = "Proceedings of the 12th International Symposium on Artificial Life and Robotis, AROB 12th'07",

pages = "597--600",

booktitle = "Proceedings of the 12th International Symposium on Artificial Life and Robotics, AROB 12th'07",

}

Zhao, XM, Cheung, YM, Chen, L & Aihara, K 2007, A new technique for adjusting the learning rate of RPEM algorithm automatically. in Proceedings of the 12th International Symposium on Artificial Life and Robotics, AROB 12th'07. Proceedings of the 12th International Symposium on Artificial Life and Robotis, AROB 12th'07, pp. 597-600, 12th International Symposium on Artificial Life and Robotics, AROB 12th'07, Oita, Japan, 25/01/07.

A new technique for adjusting the learning rate of RPEM algorithm automatically. / Zhao, Xing Ming; Cheung, Yiu Ming; Chen, Luonan et al.
Proceedings of the 12th International Symposium on Artificial Life and Robotics, AROB 12th'07. 2007. p. 597-600 (Proceedings of the 12th International Symposium on Artificial Life and Robotis, AROB 12th'07).

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

TY - GEN

T1 - A new technique for adjusting the learning rate of RPEM algorithm automatically

AU - Zhao, Xing Ming

AU - Cheung, Yiu Ming

AU - Chen, Luonan

AU - Aihara, Kazuyuki

PY - 2007

Y1 - 2007

N2 - Recently, a new Rival Penalized Expectation Maximization (RPEM) algorithm has been proposed for estimating the parameters of the normal mixture model, meanwhile determining the number of classes automatically. The RPEM is an adaptive algorithm utilizing a small constant learning rate. To speed up its convergence speed, this paper proposes a new method to dynamically adjust the learning rate of the RPEM algorithm on line. The numerical results have shown the promising results of the proposed algorithm.

AB - Recently, a new Rival Penalized Expectation Maximization (RPEM) algorithm has been proposed for estimating the parameters of the normal mixture model, meanwhile determining the number of classes automatically. The RPEM is an adaptive algorithm utilizing a small constant learning rate. To speed up its convergence speed, this paper proposes a new method to dynamically adjust the learning rate of the RPEM algorithm on line. The numerical results have shown the promising results of the proposed algorithm.

UR - http://www.scopus.com/inward/record.url?scp=78549283474&partnerID=8YFLogxK

M3 - Conference proceeding

AN - SCOPUS:78549283474

SN - 9784990288013

T3 - Proceedings of the 12th International Symposium on Artificial Life and Robotis, AROB 12th'07

SP - 597

EP - 600

BT - Proceedings of the 12th International Symposium on Artificial Life and Robotics, AROB 12th'07

T2 - 12th International Symposium on Artificial Life and Robotics, AROB 12th'07

Y2 - 25 January 2007 through 27 January 2007

ER -

A new technique for adjusting the learning rate of RPEM algorithm automatically

Abstract

Publication series

Conference

Scopus Subject Areas

Other files and links

Fingerprint

Cite this