Bagging.LMS: A bagging-based linear fusion with least-mean-square error update for regression

Yunfeng Wu; Cong Wang; S. C. Ng

doi:10.1109/TENCON.2006.343982

Bagging.LMS: A bagging-based linear fusion with least-mean-square error update for regression

Yunfeng Wu, Cong Wang, S. C. Ng

Hong Kong Metropolitan University

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

5 Citations (Scopus)

Abstract

The merits of linear decision fusion in multiple learner systems have been widely accepted, and their practical applications are rich in literature. In this paper we present a new linear decision fusion strategy named Bagging.LMS, which takes advantage of the least-mean-square (LMS) algorithm to update the fusion parameters in the Bagging ensemble systems. In the regression experiments on four synthetic and two benchmark data sets, we compared this method with the Bagging-based Simple Average and Adaptive Mixture of Experts ensemble methods. The empirical results show that the Bagging.LMS method may significantly reduce the regression errors versus the other two types of Bagging ensembles, which indicates the superiority of the suggested Bagging.LMS method.

Original language	English
Title of host publication	2006 IEEE Region 10 Conference, TENCON 2006
DOIs	https://doi.org/10.1109/TENCON.2006.343982
Publication status	Published - 2006
Event	2006 IEEE Region 10 Conference, TENCON 2006 - Hong Kong, China Duration: 14 Nov 2006 → 17 Nov 2006

Publication series

Name	IEEE Region 10 Annual International Conference, Proceedings/TENCON
ISSN (Print)	2159-3442
ISSN (Electronic)	2159-3450

Conference

Conference	2006 IEEE Region 10 Conference, TENCON 2006
Country/Territory	China
City	Hong Kong
Period	14/11/06 → 17/11/06

Access to Document

10.1109/TENCON.2006.343982

Cite this

@inproceedings{bde951e0d05d45ef97040379d1e3dee2,

title = "Bagging.LMS: A bagging-based linear fusion with least-mean-square error update for regression",

abstract = "The merits of linear decision fusion in multiple learner systems have been widely accepted, and their practical applications are rich in literature. In this paper we present a new linear decision fusion strategy named Bagging.LMS, which takes advantage of the least-mean-square (LMS) algorithm to update the fusion parameters in the Bagging ensemble systems. In the regression experiments on four synthetic and two benchmark data sets, we compared this method with the Bagging-based Simple Average and Adaptive Mixture of Experts ensemble methods. The empirical results show that the Bagging.LMS method may significantly reduce the regression errors versus the other two types of Bagging ensembles, which indicates the superiority of the suggested Bagging.LMS method.",

author = "Yunfeng Wu and Cong Wang and Ng, {S. C.}",

year = "2006",

doi = "10.1109/TENCON.2006.343982",

language = "English",

isbn = "1424405491",

series = "IEEE Region 10 Annual International Conference, Proceedings/TENCON",

booktitle = "2006 IEEE Region 10 Conference, TENCON 2006",

note = "2006 IEEE Region 10 Conference, TENCON 2006 ; Conference date: 14-11-2006 Through 17-11-2006",

}

TY - GEN

T1 - Bagging.LMS

T2 - 2006 IEEE Region 10 Conference, TENCON 2006

AU - Wu, Yunfeng

AU - Wang, Cong

AU - Ng, S. C.

PY - 2006

Y1 - 2006

N2 - The merits of linear decision fusion in multiple learner systems have been widely accepted, and their practical applications are rich in literature. In this paper we present a new linear decision fusion strategy named Bagging.LMS, which takes advantage of the least-mean-square (LMS) algorithm to update the fusion parameters in the Bagging ensemble systems. In the regression experiments on four synthetic and two benchmark data sets, we compared this method with the Bagging-based Simple Average and Adaptive Mixture of Experts ensemble methods. The empirical results show that the Bagging.LMS method may significantly reduce the regression errors versus the other two types of Bagging ensembles, which indicates the superiority of the suggested Bagging.LMS method.

AB - The merits of linear decision fusion in multiple learner systems have been widely accepted, and their practical applications are rich in literature. In this paper we present a new linear decision fusion strategy named Bagging.LMS, which takes advantage of the least-mean-square (LMS) algorithm to update the fusion parameters in the Bagging ensemble systems. In the regression experiments on four synthetic and two benchmark data sets, we compared this method with the Bagging-based Simple Average and Adaptive Mixture of Experts ensemble methods. The empirical results show that the Bagging.LMS method may significantly reduce the regression errors versus the other two types of Bagging ensembles, which indicates the superiority of the suggested Bagging.LMS method.

UR - http://www.scopus.com/inward/record.url?scp=34547591932&partnerID=8YFLogxK

U2 - 10.1109/TENCON.2006.343982

DO - 10.1109/TENCON.2006.343982

M3 - Conference contribution

AN - SCOPUS:34547591932

SN - 1424405491

SN - 9781424405497

T3 - IEEE Region 10 Annual International Conference, Proceedings/TENCON

BT - 2006 IEEE Region 10 Conference, TENCON 2006

Y2 - 14 November 2006 through 17 November 2006

ER -

Bagging.LMS: A bagging-based linear fusion with least-mean-square error update for regression

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this