On the construction of Support Wavelet Network

Junbin Gao; Daming Shi; Fei Chen

doi:10.1109/ICSMC.2004.1400833

On the construction of Support Wavelet Network

Junbin Gao, Daming Shi, Fei Chen

Research output: Book chapter/Published conference paper › Conference paper › peer-review

6 Citations (Scopus)

11 Downloads (Pure)

Abstract

Wavelet networks have emerged as a powerful tool for nonparametric estimation. It is a method implementing inverse discrete wavelet transform with coefficient optimization techniques from machine learning field. However, conventional ways to construct wavelet networks are based on empirical risk minimization (ERM) principle, which has been proven not as robust as structural risk minimization (SRM) principle. Thus, to explore the optimal architecture of wavelet networks, we constructed wavelet networks based on SRM principle. This paper describes the kernel-based way to optimize the architecture of wavelet networks. Based on the frame theory, wavelet kernel functions are found. After that, the wavelet network is constructed with support vectors generated by the wavelet kernel functions.

Original language	English
Title of host publication	2004 IEEE International Conference on Systems, Man & Cybernetics
Place of Publication	USA
Publisher	IEEE
Pages	3204-3207
Number of pages	4
Volume	4
DOIs	https://doi.org/10.1109/ICSMC.2004.1400833
Publication status	Published - 2004
Event	IEEE Conference on Systems, Man and Cybernetics - The Hague, The Netherlands, Netherlands Duration: 10 Oct 2004 → 13 Oct 2004

Conference

Conference	IEEE Conference on Systems, Man and Cybernetics
Country/Territory	Netherlands
Period	10/10/04 → 13/10/04

Access to Document

10.1109/ICSMC.2004.1400833

ManuscriptAccepted author manuscript, 250 KB

Cite this

@inproceedings{8b0423378fdc49debef9f470ede9e4b7,

title = "On the construction of Support Wavelet Network",

abstract = "Wavelet networks have emerged as a powerful tool for nonparametric estimation. It is a method implementing inverse discrete wavelet transform with coefficient optimization techniques from machine learning field. However, conventional ways to construct wavelet networks are based on empirical risk minimization (ERM) principle, which has been proven not as robust as structural risk minimization (SRM) principle. Thus, to explore the optimal architecture of wavelet networks, we constructed wavelet networks based on SRM principle. This paper describes the kernel-based way to optimize the architecture of wavelet networks. Based on the frame theory, wavelet kernel functions are found. After that, the wavelet network is constructed with support vectors generated by the wavelet kernel functions.",

keywords = "Open access version available, Neural Networks, SRM",

author = "Junbin Gao and Daming Shi and Fei Chen",

note = "Imported on 03 May 2017 - DigiTool details were: publisher = USA: IEEE, 2004. Event dates (773o) = 10-13 October 2004; Parent title (773t) = IEEE Conference on Systems, Man and Cybernetics. ISSNs: 1062-922X; ; IEEE Conference on Systems, Man and Cybernetics ; Conference date: 10-10-2004 Through 13-10-2004",

year = "2004",

doi = "10.1109/ICSMC.2004.1400833",

language = "English",

volume = "4",

pages = "3204--3207",

booktitle = "2004 IEEE International Conference on Systems, Man & Cybernetics",

publisher = "IEEE",

}

TY - GEN

T1 - On the construction of Support Wavelet Network

AU - Gao, Junbin

AU - Shi, Daming

AU - Chen, Fei

N1 - Imported on 03 May 2017 - DigiTool details were: publisher = USA: IEEE, 2004. Event dates (773o) = 10-13 October 2004; Parent title (773t) = IEEE Conference on Systems, Man and Cybernetics. ISSNs: 1062-922X;

PY - 2004

Y1 - 2004

N2 - Wavelet networks have emerged as a powerful tool for nonparametric estimation. It is a method implementing inverse discrete wavelet transform with coefficient optimization techniques from machine learning field. However, conventional ways to construct wavelet networks are based on empirical risk minimization (ERM) principle, which has been proven not as robust as structural risk minimization (SRM) principle. Thus, to explore the optimal architecture of wavelet networks, we constructed wavelet networks based on SRM principle. This paper describes the kernel-based way to optimize the architecture of wavelet networks. Based on the frame theory, wavelet kernel functions are found. After that, the wavelet network is constructed with support vectors generated by the wavelet kernel functions.

AB - Wavelet networks have emerged as a powerful tool for nonparametric estimation. It is a method implementing inverse discrete wavelet transform with coefficient optimization techniques from machine learning field. However, conventional ways to construct wavelet networks are based on empirical risk minimization (ERM) principle, which has been proven not as robust as structural risk minimization (SRM) principle. Thus, to explore the optimal architecture of wavelet networks, we constructed wavelet networks based on SRM principle. This paper describes the kernel-based way to optimize the architecture of wavelet networks. Based on the frame theory, wavelet kernel functions are found. After that, the wavelet network is constructed with support vectors generated by the wavelet kernel functions.

KW - Open access version available

KW - Neural Networks

KW - SRM

U2 - 10.1109/ICSMC.2004.1400833

DO - 10.1109/ICSMC.2004.1400833

M3 - Conference paper

VL - 4

SP - 3204

EP - 3207

BT - 2004 IEEE International Conference on Systems, Man & Cybernetics

PB - IEEE

CY - USA

T2 - IEEE Conference on Systems, Man and Cybernetics

Y2 - 10 October 2004 through 13 October 2004

ER -

On the construction of Support Wavelet Network

Abstract

Conference

Access to Document

Fingerprint

Cite this