Normalization approach to the stochastic gradient radial basis function network algorithm for odor sensing systems

Namyong Kim; Hyung Gi Byun; Krishna C. Persaud

doi:10.1016/j.snb.2007.01.001

Normalization approach to the stochastic gradient radial basis function network algorithm for odor sensing systems

Namyong Kim, Hyung Gi Byun, Krishna C. Persaud

Research output: Contribution to journal › Article › peer-review

Abstract

A method of adapting centers and weights in the radial basis function network (RBFN) is introduced using a normalization method to the stochastic gradient (RBFN-SG) algorithm for odor classification. The RBFN input data vector is from a conducting polymer sensor array. Using Taylor's expansion, a normalized form of the RBFN-SG algorithm is derived. The tracking dynamics of the normalized method appear to be less sensitive to widely varying inputs than the RBFN-SG. Experimental results of the proposed method have shown a faster learning speed, a lower mean squared error (MSE) and better classification performance. © 2007 Elsevier B.V. All rights reserved.

Original language	English
Pages (from-to)	407-412
Number of pages	5
Journal	Sensors and Actuators B: Chemical: international journal devoted to research and development of physical and chemical transducers
Volume	124
Issue number	2
DOIs	https://doi.org/10.1016/j.snb.2007.01.001
Publication status	Published - 26 Jun 2007

Keywords

Normalization
Odor
RBFN
Stochastic gradient

Access to Document

10.1016/j.snb.2007.01.001

Cite this

@article{52b7c88233f94edc8c8e1d58ae1e9c1c,

title = "Normalization approach to the stochastic gradient radial basis function network algorithm for odor sensing systems",

abstract = "A method of adapting centers and weights in the radial basis function network (RBFN) is introduced using a normalization method to the stochastic gradient (RBFN-SG) algorithm for odor classification. The RBFN input data vector is from a conducting polymer sensor array. Using Taylor's expansion, a normalized form of the RBFN-SG algorithm is derived. The tracking dynamics of the normalized method appear to be less sensitive to widely varying inputs than the RBFN-SG. Experimental results of the proposed method have shown a faster learning speed, a lower mean squared error (MSE) and better classification performance. {\textcopyright} 2007 Elsevier B.V. All rights reserved.",

keywords = "Normalization, Odor, RBFN, Stochastic gradient",

author = "Namyong Kim and Byun, {Hyung Gi} and Persaud, {Krishna C.}",

note = "Times Cited: 1",

year = "2007",

month = jun,

day = "26",

doi = "10.1016/j.snb.2007.01.001",

language = "English",

volume = "124",

pages = "407--412",

journal = "Sensors and Actuators B: Chemical: international journal devoted to research and development of physical and chemical transducers",

issn = "0925-4005",

publisher = "Elsevier BV",

number = "2",

}

Normalization approach to the stochastic gradient radial basis function network algorithm for odor sensing systems. / Kim, Namyong; Byun, Hyung Gi; Persaud, Krishna C.
In: Sensors and Actuators B: Chemical: international journal devoted to research and development of physical and chemical transducers, Vol. 124, No. 2, 26.06.2007, p. 407-412.

Research output: Contribution to journal › Article › peer-review

TY - JOUR

T1 - Normalization approach to the stochastic gradient radial basis function network algorithm for odor sensing systems

AU - Kim, Namyong

AU - Byun, Hyung Gi

AU - Persaud, Krishna C.

N1 - Times Cited: 1

PY - 2007/6/26

Y1 - 2007/6/26

N2 - A method of adapting centers and weights in the radial basis function network (RBFN) is introduced using a normalization method to the stochastic gradient (RBFN-SG) algorithm for odor classification. The RBFN input data vector is from a conducting polymer sensor array. Using Taylor's expansion, a normalized form of the RBFN-SG algorithm is derived. The tracking dynamics of the normalized method appear to be less sensitive to widely varying inputs than the RBFN-SG. Experimental results of the proposed method have shown a faster learning speed, a lower mean squared error (MSE) and better classification performance. © 2007 Elsevier B.V. All rights reserved.

AB - A method of adapting centers and weights in the radial basis function network (RBFN) is introduced using a normalization method to the stochastic gradient (RBFN-SG) algorithm for odor classification. The RBFN input data vector is from a conducting polymer sensor array. Using Taylor's expansion, a normalized form of the RBFN-SG algorithm is derived. The tracking dynamics of the normalized method appear to be less sensitive to widely varying inputs than the RBFN-SG. Experimental results of the proposed method have shown a faster learning speed, a lower mean squared error (MSE) and better classification performance. © 2007 Elsevier B.V. All rights reserved.

KW - Normalization

KW - Odor

KW - RBFN

KW - Stochastic gradient

U2 - 10.1016/j.snb.2007.01.001

DO - 10.1016/j.snb.2007.01.001

M3 - Article

SN - 0925-4005

VL - 124

SP - 407

EP - 412

JO - Sensors and Actuators B: Chemical: international journal devoted to research and development of physical and chemical transducers

JF - Sensors and Actuators B: Chemical: international journal devoted to research and development of physical and chemical transducers

IS - 2

ER -

Normalization approach to the stochastic gradient radial basis function network algorithm for odor sensing systems

Abstract

Keywords

Access to Document

Fingerprint

Cite this