Receding Horizon Prediction by Bayesian Combination of Multiple Predictors

Fredrik Ståhl; Rolf Johansson

Receding Horizon Prediction by Bayesian Combination of Multiple Predictors

Research output: Chapter in Book/Report/Conference proceeding › Paper in conference proceeding › peer-review

259 Downloads (Pure)

Abstract

This paper presents a novel online approach of merging multiple different predictors of time-varying dynamics into a single optimized prediction. Different predictors are merged by recursive weighting into a single prediction using regularized optimization. The approach is evaluated on two different cases of data with shifting dynamics; one example of prediction using several approximate models of a linear system and one case of glucose prediction on a non-linear physiologically based simulated type I diabetes data using several parallel linear predictors. The performance of the combined prediction significantly reduced the total prediction error compared to each
predictor in each example.

Original language	English
Title of host publication	Proc. 51st IEEE Conf. Decision and Control (CDC 2012), December 10-13, 2012. Maui, Hawaii, USA
Publisher	IEEE - Institute of Electrical and Electronics Engineers Inc.
Pages	5278-5285
Number of pages	8
Publication status	Published - 2012
Event	51st IEEE Conference on Decision and Control, 2012 - Maui, Hawaii, Maui, Hawaii, United States Duration: 2012 Dec 10 → 2012 Dec 13 Conference number: 51 http://www.ieeecss.org/CAB/conferences/cdc2012/

Publication series

Name
ISSN (Print)	0191-2216

Conference

Conference	51st IEEE Conference on Decision and Control, 2012
Abbreviated title	cdc2012
Country/Territory	United States
City	Maui, Hawaii
Period	2012/12/10 → 2012/12/13
Internet address	http://www.ieeecss.org/CAB/conferences/cdc2012/

Subject classification (UKÄ)

Control Engineering

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

CDC2012_FS

Cite this

@inproceedings{6ed27d9878f04dee9b55d89e585c99f8,

title = "Receding Horizon Prediction by Bayesian Combination of Multiple Predictors",

abstract = "This paper presents a novel online approach of merging multiple different predictors of time-varying dynamics into a single optimized prediction. Different predictors are merged by recursive weighting into a single prediction using regularized optimization. The approach is evaluated on two different cases of data with shifting dynamics; one example of prediction using several approximate models of a linear system and one case of glucose prediction on a non-linear physiologically based simulated type I diabetes data using several parallel linear predictors. The performance of the combined prediction significantly reduced the total prediction error compared to each predictor in each example.",

author = "Fredrik St{\aa}hl and Rolf Johansson",

year = "2012",

language = "English",

publisher = "IEEE - Institute of Electrical and Electronics Engineers Inc.",

pages = "5278--5285",

booktitle = "Proc. 51st IEEE Conf. Decision and Control (CDC 2012), December 10-13, 2012. Maui, Hawaii, USA",

address = "United States",

note = "51st IEEE Conference on Decision and Control, 2012, cdc2012 ; Conference date: 10-12-2012 Through 13-12-2012",

url = "http://www.ieeecss.org/CAB/conferences/cdc2012/",

}

Ståhl, F & Johansson, R 2012, Receding Horizon Prediction by Bayesian Combination of Multiple Predictors. in Proc. 51st IEEE Conf. Decision and Control (CDC 2012), December 10-13, 2012. Maui, Hawaii, USA. IEEE - Institute of Electrical and Electronics Engineers Inc., pp. 5278-5285, 51st IEEE Conference on Decision and Control, 2012, Maui, Hawaii, United States, 2012/12/10.

Receding Horizon Prediction by Bayesian Combination of Multiple Predictors. / Ståhl, Fredrik; Johansson, Rolf.
Proc. 51st IEEE Conf. Decision and Control (CDC 2012), December 10-13, 2012. Maui, Hawaii, USA. IEEE - Institute of Electrical and Electronics Engineers Inc., 2012. p. 5278-5285.

Research output: Chapter in Book/Report/Conference proceeding › Paper in conference proceeding › peer-review

TY - GEN

T1 - Receding Horizon Prediction by Bayesian Combination of Multiple Predictors

AU - Ståhl, Fredrik

AU - Johansson, Rolf

N1 - Conference code: 51

PY - 2012

Y1 - 2012

N2 - This paper presents a novel online approach of merging multiple different predictors of time-varying dynamics into a single optimized prediction. Different predictors are merged by recursive weighting into a single prediction using regularized optimization. The approach is evaluated on two different cases of data with shifting dynamics; one example of prediction using several approximate models of a linear system and one case of glucose prediction on a non-linear physiologically based simulated type I diabetes data using several parallel linear predictors. The performance of the combined prediction significantly reduced the total prediction error compared to each predictor in each example.

AB - This paper presents a novel online approach of merging multiple different predictors of time-varying dynamics into a single optimized prediction. Different predictors are merged by recursive weighting into a single prediction using regularized optimization. The approach is evaluated on two different cases of data with shifting dynamics; one example of prediction using several approximate models of a linear system and one case of glucose prediction on a non-linear physiologically based simulated type I diabetes data using several parallel linear predictors. The performance of the combined prediction significantly reduced the total prediction error compared to each predictor in each example.

UR - https://www.scopus.com/pages/publications/84874258116

M3 - Paper in conference proceeding

SP - 5278

EP - 5285

BT - Proc. 51st IEEE Conf. Decision and Control (CDC 2012), December 10-13, 2012. Maui, Hawaii, USA

PB - IEEE - Institute of Electrical and Electronics Engineers Inc.

T2 - 51st IEEE Conference on Decision and Control, 2012

Y2 - 10 December 2012 through 13 December 2012

ER -

Receding Horizon Prediction by Bayesian Combination of Multiple Predictors

Abstract

Publication series

Conference

Subject classification (UKÄ)

UN SDGs

Access to Document

Other files and links

Fingerprint

DIAdvisor

Cite this

Receding Horizon Prediction by Bayesian Combination of Multiple Predictors

Abstract

Publication series

Conference

Subject classification (UKÄ)

UN SDGs

Access to Document

Other files and links

Fingerprint

Projects

DIAdvisor

Cite this