Temporal sequence learning and recognition with dynamic SOM

Qiong Liu, Sylvian Ray, Stephen E Levinson, Thomas S Huang, Jun Huang

Research output: Contribution to conferencePaper

Abstract

The purpose of this paper is to propose a map-like artificial neural network for temporal sequence pattern clustering. The map construction in our presentation is related to the Self-Organizing Map (SOM) idea. SOM idea was originally designed for static pattern learning and recognition. It has been found efficient for organizing high dimensional data sets. One of the biggest limitations of traditional SOM technique is caused by its static characteristics. In this paper, we proposed a new neural network construction model and its corresponding training algorithm based on traditional SOM training technology and Back-Propagation training technology. It overcomes the static limitation of traditional SOM, and tries to reach a new stage for dynamic pattern clustering, and recognition. At the end of this paper, we will give some experimental results for testing this proposed method on real speech data.

Original languageEnglish (US)
Pages2970-2975
Number of pages6
StatePublished - Dec 1 1999
EventInternational Joint Conference on Neural Networks (IJCNN'99) - Washington, DC, USA
Duration: Jul 10 1999Jul 16 1999

Other

OtherInternational Joint Conference on Neural Networks (IJCNN'99)
CityWashington, DC, USA
Period7/10/997/16/99

Fingerprint

Self organizing maps
Neural networks
Backpropagation
Testing

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence

Cite this

Liu, Q., Ray, S., Levinson, S. E., Huang, T. S., & Huang, J. (1999). Temporal sequence learning and recognition with dynamic SOM. 2970-2975. Paper presented at International Joint Conference on Neural Networks (IJCNN'99), Washington, DC, USA, .

Temporal sequence learning and recognition with dynamic SOM. / Liu, Qiong; Ray, Sylvian; Levinson, Stephen E; Huang, Thomas S; Huang, Jun.

1999. 2970-2975 Paper presented at International Joint Conference on Neural Networks (IJCNN'99), Washington, DC, USA, .

Research output: Contribution to conferencePaper

Liu, Q, Ray, S, Levinson, SE, Huang, TS & Huang, J 1999, 'Temporal sequence learning and recognition with dynamic SOM' Paper presented at International Joint Conference on Neural Networks (IJCNN'99), Washington, DC, USA, 7/10/99 - 7/16/99, pp. 2970-2975.
Liu Q, Ray S, Levinson SE, Huang TS, Huang J. Temporal sequence learning and recognition with dynamic SOM. 1999. Paper presented at International Joint Conference on Neural Networks (IJCNN'99), Washington, DC, USA, .
Liu, Qiong ; Ray, Sylvian ; Levinson, Stephen E ; Huang, Thomas S ; Huang, Jun. / Temporal sequence learning and recognition with dynamic SOM. Paper presented at International Joint Conference on Neural Networks (IJCNN'99), Washington, DC, USA, .6 p.
@conference{e7f108f4f26d40d1b67d0de9d302485c,
title = "Temporal sequence learning and recognition with dynamic SOM",
abstract = "The purpose of this paper is to propose a map-like artificial neural network for temporal sequence pattern clustering. The map construction in our presentation is related to the Self-Organizing Map (SOM) idea. SOM idea was originally designed for static pattern learning and recognition. It has been found efficient for organizing high dimensional data sets. One of the biggest limitations of traditional SOM technique is caused by its static characteristics. In this paper, we proposed a new neural network construction model and its corresponding training algorithm based on traditional SOM training technology and Back-Propagation training technology. It overcomes the static limitation of traditional SOM, and tries to reach a new stage for dynamic pattern clustering, and recognition. At the end of this paper, we will give some experimental results for testing this proposed method on real speech data.",
author = "Qiong Liu and Sylvian Ray and Levinson, {Stephen E} and Huang, {Thomas S} and Jun Huang",
year = "1999",
month = "12",
day = "1",
language = "English (US)",
pages = "2970--2975",
note = "International Joint Conference on Neural Networks (IJCNN'99) ; Conference date: 10-07-1999 Through 16-07-1999",

}

TY - CONF

T1 - Temporal sequence learning and recognition with dynamic SOM

AU - Liu, Qiong

AU - Ray, Sylvian

AU - Levinson, Stephen E

AU - Huang, Thomas S

AU - Huang, Jun

PY - 1999/12/1

Y1 - 1999/12/1

N2 - The purpose of this paper is to propose a map-like artificial neural network for temporal sequence pattern clustering. The map construction in our presentation is related to the Self-Organizing Map (SOM) idea. SOM idea was originally designed for static pattern learning and recognition. It has been found efficient for organizing high dimensional data sets. One of the biggest limitations of traditional SOM technique is caused by its static characteristics. In this paper, we proposed a new neural network construction model and its corresponding training algorithm based on traditional SOM training technology and Back-Propagation training technology. It overcomes the static limitation of traditional SOM, and tries to reach a new stage for dynamic pattern clustering, and recognition. At the end of this paper, we will give some experimental results for testing this proposed method on real speech data.

AB - The purpose of this paper is to propose a map-like artificial neural network for temporal sequence pattern clustering. The map construction in our presentation is related to the Self-Organizing Map (SOM) idea. SOM idea was originally designed for static pattern learning and recognition. It has been found efficient for organizing high dimensional data sets. One of the biggest limitations of traditional SOM technique is caused by its static characteristics. In this paper, we proposed a new neural network construction model and its corresponding training algorithm based on traditional SOM training technology and Back-Propagation training technology. It overcomes the static limitation of traditional SOM, and tries to reach a new stage for dynamic pattern clustering, and recognition. At the end of this paper, we will give some experimental results for testing this proposed method on real speech data.

UR - http://www.scopus.com/inward/record.url?scp=0033326226&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0033326226&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0033326226

SP - 2970

EP - 2975

ER -