Training neural pattern classifiers with a mean field theory learning algorithm

Jung-Hsien Chiang, Russell Pimmel

Research output: Contribution to conferencePaper

Abstract

This study evaluated the mean field theory (MFT) learning algorithm with pattern classifiers and compared its performance to that of the back-propagation (BP) learning algorithm. The deterministic MFT learning algorithm replaced the time-consuming stochastic Boltzmann machine learning algorithm. In our simulation studies, the MFT was superior to the BP learning algorithm with two nonlinearly separable binary mappings, with the Iris data classification problem, and with a simulated signal classification problem. The MFF algorithm always required a substantially smaller number of training epochs than the BP algorithm, and in the case of the signal classification problem the MFT algorithm achieved much lower errors and misclassification rates.

Original languageEnglish
Pages285-290
Number of pages6
Publication statusPublished - 1992 Dec 1
EventProceedings of the 1992 Artificial Neural Networks in Engineering, ANNIE'92 - St.Louis, MO, USA
Duration: 1992 Nov 151992 Nov 18

Other

OtherProceedings of the 1992 Artificial Neural Networks in Engineering, ANNIE'92
CitySt.Louis, MO, USA
Period92-11-1592-11-18

Fingerprint

Mean field theory
Learning algorithms
Classifiers
Backpropagation
Backpropagation algorithms
Learning systems

All Science Journal Classification (ASJC) codes

  • Software

Cite this

Chiang, J-H., & Pimmel, R. (1992). Training neural pattern classifiers with a mean field theory learning algorithm. 285-290. Paper presented at Proceedings of the 1992 Artificial Neural Networks in Engineering, ANNIE'92, St.Louis, MO, USA, .
Chiang, Jung-Hsien ; Pimmel, Russell. / Training neural pattern classifiers with a mean field theory learning algorithm. Paper presented at Proceedings of the 1992 Artificial Neural Networks in Engineering, ANNIE'92, St.Louis, MO, USA, .6 p.
@conference{9a42e8bd6d8441cbadf7e9a06ff4f5d6,
title = "Training neural pattern classifiers with a mean field theory learning algorithm",
abstract = "This study evaluated the mean field theory (MFT) learning algorithm with pattern classifiers and compared its performance to that of the back-propagation (BP) learning algorithm. The deterministic MFT learning algorithm replaced the time-consuming stochastic Boltzmann machine learning algorithm. In our simulation studies, the MFT was superior to the BP learning algorithm with two nonlinearly separable binary mappings, with the Iris data classification problem, and with a simulated signal classification problem. The MFF algorithm always required a substantially smaller number of training epochs than the BP algorithm, and in the case of the signal classification problem the MFT algorithm achieved much lower errors and misclassification rates.",
author = "Jung-Hsien Chiang and Russell Pimmel",
year = "1992",
month = "12",
day = "1",
language = "English",
pages = "285--290",
note = "Proceedings of the 1992 Artificial Neural Networks in Engineering, ANNIE'92 ; Conference date: 15-11-1992 Through 18-11-1992",

}

Chiang, J-H & Pimmel, R 1992, 'Training neural pattern classifiers with a mean field theory learning algorithm' Paper presented at Proceedings of the 1992 Artificial Neural Networks in Engineering, ANNIE'92, St.Louis, MO, USA, 92-11-15 - 92-11-18, pp. 285-290.

Training neural pattern classifiers with a mean field theory learning algorithm. / Chiang, Jung-Hsien; Pimmel, Russell.

1992. 285-290 Paper presented at Proceedings of the 1992 Artificial Neural Networks in Engineering, ANNIE'92, St.Louis, MO, USA, .

Research output: Contribution to conferencePaper

TY - CONF

T1 - Training neural pattern classifiers with a mean field theory learning algorithm

AU - Chiang, Jung-Hsien

AU - Pimmel, Russell

PY - 1992/12/1

Y1 - 1992/12/1

N2 - This study evaluated the mean field theory (MFT) learning algorithm with pattern classifiers and compared its performance to that of the back-propagation (BP) learning algorithm. The deterministic MFT learning algorithm replaced the time-consuming stochastic Boltzmann machine learning algorithm. In our simulation studies, the MFT was superior to the BP learning algorithm with two nonlinearly separable binary mappings, with the Iris data classification problem, and with a simulated signal classification problem. The MFF algorithm always required a substantially smaller number of training epochs than the BP algorithm, and in the case of the signal classification problem the MFT algorithm achieved much lower errors and misclassification rates.

AB - This study evaluated the mean field theory (MFT) learning algorithm with pattern classifiers and compared its performance to that of the back-propagation (BP) learning algorithm. The deterministic MFT learning algorithm replaced the time-consuming stochastic Boltzmann machine learning algorithm. In our simulation studies, the MFT was superior to the BP learning algorithm with two nonlinearly separable binary mappings, with the Iris data classification problem, and with a simulated signal classification problem. The MFF algorithm always required a substantially smaller number of training epochs than the BP algorithm, and in the case of the signal classification problem the MFT algorithm achieved much lower errors and misclassification rates.

UR - http://www.scopus.com/inward/record.url?scp=0027105487&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0027105487&partnerID=8YFLogxK

M3 - Paper

SP - 285

EP - 290

ER -

Chiang J-H, Pimmel R. Training neural pattern classifiers with a mean field theory learning algorithm. 1992. Paper presented at Proceedings of the 1992 Artificial Neural Networks in Engineering, ANNIE'92, St.Louis, MO, USA, .