Gaussian process regression using Laplace approximations under localization uncertainty

Mahdi Jadaliha, Yunfei Xu, Jongeun Choi

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

5 Citations (Scopus)

Abstract

In this paper, we formulate Gaussian process regression with observations under the localization uncertainty. In our formulation, effects of observations, measurement noise, localization uncertainty and prior distributions are all correctly incorporated in the posterior predictive statistics. The analytically intractable posterior predictive statistics are proposed to be approximated by Laplace approximations in different degrees of complexity. Such approximations have been carefully tailored to our problems and their approximation errors and complexity are analyzed. Simulation results demonstrate that the proposed approaches perform much better than approaches without considering the localization uncertainty correctly.

Original languageEnglish
Title of host publication2012 American Control Conference, ACC 2012
Pages1394-1399
Number of pages6
Publication statusPublished - 2012 Nov 26
Event2012 American Control Conference, ACC 2012 - Montreal, QC, Canada
Duration: 2012 Jun 272012 Jun 29

Other

Other2012 American Control Conference, ACC 2012
CountryCanada
CityMontreal, QC
Period12/6/2712/6/29

Fingerprint

Statistics
Uncertainty

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering

Cite this

Jadaliha, M., Xu, Y., & Choi, J. (2012). Gaussian process regression using Laplace approximations under localization uncertainty. In 2012 American Control Conference, ACC 2012 (pp. 1394-1399). [6314793]
Jadaliha, Mahdi ; Xu, Yunfei ; Choi, Jongeun. / Gaussian process regression using Laplace approximations under localization uncertainty. 2012 American Control Conference, ACC 2012. 2012. pp. 1394-1399
@inproceedings{8bea508eeddb41e6b01e0b730b693c39,
title = "Gaussian process regression using Laplace approximations under localization uncertainty",
abstract = "In this paper, we formulate Gaussian process regression with observations under the localization uncertainty. In our formulation, effects of observations, measurement noise, localization uncertainty and prior distributions are all correctly incorporated in the posterior predictive statistics. The analytically intractable posterior predictive statistics are proposed to be approximated by Laplace approximations in different degrees of complexity. Such approximations have been carefully tailored to our problems and their approximation errors and complexity are analyzed. Simulation results demonstrate that the proposed approaches perform much better than approaches without considering the localization uncertainty correctly.",
author = "Mahdi Jadaliha and Yunfei Xu and Jongeun Choi",
year = "2012",
month = "11",
day = "26",
language = "English",
isbn = "9781457710957",
pages = "1394--1399",
booktitle = "2012 American Control Conference, ACC 2012",

}

Jadaliha, M, Xu, Y & Choi, J 2012, Gaussian process regression using Laplace approximations under localization uncertainty. in 2012 American Control Conference, ACC 2012., 6314793, pp. 1394-1399, 2012 American Control Conference, ACC 2012, Montreal, QC, Canada, 12/6/27.

Gaussian process regression using Laplace approximations under localization uncertainty. / Jadaliha, Mahdi; Xu, Yunfei; Choi, Jongeun.

2012 American Control Conference, ACC 2012. 2012. p. 1394-1399 6314793.

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

TY - GEN

T1 - Gaussian process regression using Laplace approximations under localization uncertainty

AU - Jadaliha, Mahdi

AU - Xu, Yunfei

AU - Choi, Jongeun

PY - 2012/11/26

Y1 - 2012/11/26

N2 - In this paper, we formulate Gaussian process regression with observations under the localization uncertainty. In our formulation, effects of observations, measurement noise, localization uncertainty and prior distributions are all correctly incorporated in the posterior predictive statistics. The analytically intractable posterior predictive statistics are proposed to be approximated by Laplace approximations in different degrees of complexity. Such approximations have been carefully tailored to our problems and their approximation errors and complexity are analyzed. Simulation results demonstrate that the proposed approaches perform much better than approaches without considering the localization uncertainty correctly.

AB - In this paper, we formulate Gaussian process regression with observations under the localization uncertainty. In our formulation, effects of observations, measurement noise, localization uncertainty and prior distributions are all correctly incorporated in the posterior predictive statistics. The analytically intractable posterior predictive statistics are proposed to be approximated by Laplace approximations in different degrees of complexity. Such approximations have been carefully tailored to our problems and their approximation errors and complexity are analyzed. Simulation results demonstrate that the proposed approaches perform much better than approaches without considering the localization uncertainty correctly.

UR - http://www.scopus.com/inward/record.url?scp=84869440926&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84869440926&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9781457710957

SP - 1394

EP - 1399

BT - 2012 American Control Conference, ACC 2012

ER -

Jadaliha M, Xu Y, Choi J. Gaussian process regression using Laplace approximations under localization uncertainty. In 2012 American Control Conference, ACC 2012. 2012. p. 1394-1399. 6314793