An adaptive approximation for Gaussian wavelet kernel

Young Mok Ha, Ji Won Yoon

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Kernel machine plays a critical role in science community since temporal data become more important and popular with rapidly increasing big data analysis. A major problem for the machine is difficulty in constructing kernel function. We show that it is possible to adaptively estimate the parameters of Gaussian wavelet kernel in Laplace method. Our approach is constructed on an obvious fact that the gradient of the kernel with respect to a central variable of feature space becomes zero. It is remarkable that the complexity of our estimation method is O(N) for N data. In order to validate the performance of the proposed approach, we simulate two kernel regression models which exploit the proposed approach on real electricity load data from Korea power exchange and electricity consumption data from Ireland's Commission for Energy Regulation.

Original languageEnglish
Title of host publicationInternational Conference on Advanced Communication Technology, ICACT
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages576-580
Number of pages5
Volume2016-March
ISBN (Print)9788996865063
DOIs
Publication statusPublished - 2016 Mar 1
Event18th International Conference on Advanced Communications Technology, ICACT 2016 - Pyeongchang, Korea, Republic of
Duration: 2016 Jan 312016 Feb 3

Other

Other18th International Conference on Advanced Communications Technology, ICACT 2016
CountryKorea, Republic of
CityPyeongchang
Period16/1/3116/2/3

Fingerprint

Electricity
Big data

Keywords

  • adaptive inference
  • electricity load/consumption forecast
  • Gaussian wavelet kernel
  • Laplace approximation

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Cite this

Ha, Y. M., & Yoon, J. W. (2016). An adaptive approximation for Gaussian wavelet kernel. In International Conference on Advanced Communication Technology, ICACT (Vol. 2016-March, pp. 576-580). [7423478] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICACT.2016.7423478

An adaptive approximation for Gaussian wavelet kernel. / Ha, Young Mok; Yoon, Ji Won.

International Conference on Advanced Communication Technology, ICACT. Vol. 2016-March Institute of Electrical and Electronics Engineers Inc., 2016. p. 576-580 7423478.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Ha, YM & Yoon, JW 2016, An adaptive approximation for Gaussian wavelet kernel. in International Conference on Advanced Communication Technology, ICACT. vol. 2016-March, 7423478, Institute of Electrical and Electronics Engineers Inc., pp. 576-580, 18th International Conference on Advanced Communications Technology, ICACT 2016, Pyeongchang, Korea, Republic of, 16/1/31. https://doi.org/10.1109/ICACT.2016.7423478
Ha YM, Yoon JW. An adaptive approximation for Gaussian wavelet kernel. In International Conference on Advanced Communication Technology, ICACT. Vol. 2016-March. Institute of Electrical and Electronics Engineers Inc. 2016. p. 576-580. 7423478 https://doi.org/10.1109/ICACT.2016.7423478
Ha, Young Mok ; Yoon, Ji Won. / An adaptive approximation for Gaussian wavelet kernel. International Conference on Advanced Communication Technology, ICACT. Vol. 2016-March Institute of Electrical and Electronics Engineers Inc., 2016. pp. 576-580
@inproceedings{16a69daf5ad34a68a316f85036429a19,
title = "An adaptive approximation for Gaussian wavelet kernel",
abstract = "Kernel machine plays a critical role in science community since temporal data become more important and popular with rapidly increasing big data analysis. A major problem for the machine is difficulty in constructing kernel function. We show that it is possible to adaptively estimate the parameters of Gaussian wavelet kernel in Laplace method. Our approach is constructed on an obvious fact that the gradient of the kernel with respect to a central variable of feature space becomes zero. It is remarkable that the complexity of our estimation method is O(N) for N data. In order to validate the performance of the proposed approach, we simulate two kernel regression models which exploit the proposed approach on real electricity load data from Korea power exchange and electricity consumption data from Ireland's Commission for Energy Regulation.",
keywords = "adaptive inference, electricity load/consumption forecast, Gaussian wavelet kernel, Laplace approximation",
author = "Ha, {Young Mok} and Yoon, {Ji Won}",
year = "2016",
month = "3",
day = "1",
doi = "10.1109/ICACT.2016.7423478",
language = "English",
isbn = "9788996865063",
volume = "2016-March",
pages = "576--580",
booktitle = "International Conference on Advanced Communication Technology, ICACT",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - An adaptive approximation for Gaussian wavelet kernel

AU - Ha, Young Mok

AU - Yoon, Ji Won

PY - 2016/3/1

Y1 - 2016/3/1

N2 - Kernel machine plays a critical role in science community since temporal data become more important and popular with rapidly increasing big data analysis. A major problem for the machine is difficulty in constructing kernel function. We show that it is possible to adaptively estimate the parameters of Gaussian wavelet kernel in Laplace method. Our approach is constructed on an obvious fact that the gradient of the kernel with respect to a central variable of feature space becomes zero. It is remarkable that the complexity of our estimation method is O(N) for N data. In order to validate the performance of the proposed approach, we simulate two kernel regression models which exploit the proposed approach on real electricity load data from Korea power exchange and electricity consumption data from Ireland's Commission for Energy Regulation.

AB - Kernel machine plays a critical role in science community since temporal data become more important and popular with rapidly increasing big data analysis. A major problem for the machine is difficulty in constructing kernel function. We show that it is possible to adaptively estimate the parameters of Gaussian wavelet kernel in Laplace method. Our approach is constructed on an obvious fact that the gradient of the kernel with respect to a central variable of feature space becomes zero. It is remarkable that the complexity of our estimation method is O(N) for N data. In order to validate the performance of the proposed approach, we simulate two kernel regression models which exploit the proposed approach on real electricity load data from Korea power exchange and electricity consumption data from Ireland's Commission for Energy Regulation.

KW - adaptive inference

KW - electricity load/consumption forecast

KW - Gaussian wavelet kernel

KW - Laplace approximation

UR - http://www.scopus.com/inward/record.url?scp=84962786748&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84962786748&partnerID=8YFLogxK

U2 - 10.1109/ICACT.2016.7423478

DO - 10.1109/ICACT.2016.7423478

M3 - Conference contribution

SN - 9788996865063

VL - 2016-March

SP - 576

EP - 580

BT - International Conference on Advanced Communication Technology, ICACT

PB - Institute of Electrical and Electronics Engineers Inc.

ER -