Time-series Gaussian process regression based on toeplitz computation of O(N2) operations and O(N)-level storage

Y. Zhang; W.E. Leithead; D.J. Leith

doi:10.1109/CDC.2005.1582739

Time-series Gaussian process regression based on toeplitz computation of O(N2) operations and O(N)-level storage

Y. Zhang, W.E. Leithead, D.J. Leith

Electronic And Electrical Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution book

92 Citations (Scopus)

Abstract

Gaussian process (GP) regression is a Bayesian nonparametric model showing good performance in various applications. However, its hyperparameter-estimating procedure may contain numerous matrix manipulations of O(N3) arithmetic operations, in addition to the O(N2)-level storage. Motivated by handling the real-world large dataset of 24000 wind-turbine data, we propose in this paper an efficient and economical Toeplitz-computation scheme for time-series Gaussian process regression. The scheme is of O(N2) operations and O(N)-level memory requirement. Numerical experiments substantiate the effectiveness and possibility of using this Toeplitz computation for very large datasets regression (such as, containing 10000~100000 data points).

Original language	English
Title of host publication	Proceedings of the 44th IEEE Conference on Decision and Control, 2005 and 2005 European Control Conference. CDC-ECC '05
Subtitle of host publication	CDC-ECC '05
Publisher	IEEE
Pages	3711-3716
Number of pages	6
ISBN (Print)	0-7803-9567-0
DOIs	https://doi.org/10.1109/CDC.2005.1582739
Publication status	Published - Dec 2005
Event	44th IEEE Conference on Decision and Control and European Control Conference - Seville, Spain Duration: 12 Dec 2005 → 15 Dec 2005

Conference

Conference	44th IEEE Conference on Decision and Control and European Control Conference
Abbreviated title	CDC-ECC
Country/Territory	Spain
City	Seville
Period	12/12/05 → 15/12/05

Keywords

acceleration
aerodynamics
bayesian methods
noise measurement
predictive models
velocity measurement

Access to Document

10.1109/CDC.2005.1582739

Cite this

@inproceedings{8f68caa75e1d4db6834dfdcf3c91175a,

title = "Time-series Gaussian process regression based on toeplitz computation of O(N2) operations and O(N)-level storage",

abstract = "Gaussian process (GP) regression is a Bayesian nonparametric model showing good performance in various applications. However, its hyperparameter-estimating procedure may contain numerous matrix manipulations of O(N3) arithmetic operations, in addition to the O(N2)-level storage. Motivated by handling the real-world large dataset of 24000 wind-turbine data, we propose in this paper an efficient and economical Toeplitz-computation scheme for time-series Gaussian process regression. The scheme is of O(N2) operations and O(N)-level memory requirement. Numerical experiments substantiate the effectiveness and possibility of using this Toeplitz computation for very large datasets regression (such as, containing 10000~100000 data points).",

keywords = "acceleration, aerodynamics , bayesian methods, noise measurement, predictive models, velocity measurement",

author = "Y. Zhang and W.E. Leithead and D.J. Leith",

year = "2005",

month = dec,

doi = "10.1109/CDC.2005.1582739",

language = "English",

isbn = "0-7803-9567-0",

pages = "3711--3716",

booktitle = "Proceedings of the 44th IEEE Conference on Decision and Control, 2005 and 2005 European Control Conference. CDC-ECC '05",

publisher = "IEEE",

note = "44th IEEE Conference on Decision and Control and European Control Conference , CDC-ECC ; Conference date: 12-12-2005 Through 15-12-2005",

}

Zhang, Y, Leithead, WE & Leith, DJ 2005, Time-series Gaussian process regression based on toeplitz computation of O(N2) operations and O(N)-level storage. in Proceedings of the 44th IEEE Conference on Decision and Control, 2005 and 2005 European Control Conference. CDC-ECC '05: CDC-ECC '05. IEEE, pp. 3711-3716, 44th IEEE Conference on Decision and Control and European Control Conference , Seville, Spain, 12/12/05. https://doi.org/10.1109/CDC.2005.1582739

Time-series Gaussian process regression based on toeplitz computation of O(N2) operations and O(N)-level storage. / Zhang, Y.; Leithead, W.E.; Leith, D.J.
Proceedings of the 44th IEEE Conference on Decision and Control, 2005 and 2005 European Control Conference. CDC-ECC '05: CDC-ECC '05. IEEE, 2005. p. 3711-3716.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution book

TY - GEN

T1 - Time-series Gaussian process regression based on toeplitz computation of O(N2) operations and O(N)-level storage

AU - Zhang, Y.

AU - Leithead, W.E.

AU - Leith, D.J.

PY - 2005/12

Y1 - 2005/12

N2 - Gaussian process (GP) regression is a Bayesian nonparametric model showing good performance in various applications. However, its hyperparameter-estimating procedure may contain numerous matrix manipulations of O(N3) arithmetic operations, in addition to the O(N2)-level storage. Motivated by handling the real-world large dataset of 24000 wind-turbine data, we propose in this paper an efficient and economical Toeplitz-computation scheme for time-series Gaussian process regression. The scheme is of O(N2) operations and O(N)-level memory requirement. Numerical experiments substantiate the effectiveness and possibility of using this Toeplitz computation for very large datasets regression (such as, containing 10000~100000 data points).

AB - Gaussian process (GP) regression is a Bayesian nonparametric model showing good performance in various applications. However, its hyperparameter-estimating procedure may contain numerous matrix manipulations of O(N3) arithmetic operations, in addition to the O(N2)-level storage. Motivated by handling the real-world large dataset of 24000 wind-turbine data, we propose in this paper an efficient and economical Toeplitz-computation scheme for time-series Gaussian process regression. The scheme is of O(N2) operations and O(N)-level memory requirement. Numerical experiments substantiate the effectiveness and possibility of using this Toeplitz computation for very large datasets regression (such as, containing 10000~100000 data points).

KW - acceleration

KW - aerodynamics

KW - bayesian methods

KW - noise measurement

KW - predictive models

KW - velocity measurement

U2 - 10.1109/CDC.2005.1582739

DO - 10.1109/CDC.2005.1582739

M3 - Conference contribution book

SN - 0-7803-9567-0

SP - 3711

EP - 3716

BT - Proceedings of the 44th IEEE Conference on Decision and Control, 2005 and 2005 European Control Conference. CDC-ECC '05

PB - IEEE

T2 - 44th IEEE Conference on Decision and Control and European Control Conference

Y2 - 12 December 2005 through 15 December 2005

ER -

Time-series Gaussian process regression based on toeplitz computation of O(N2) operations and O(N)-level storage

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this