Multiple model regression estimation

Vladimir Cherkassky; Yunqian Ma

doi:10.1109/TNN.2005.849836

Multiple model regression estimation

Vladimir Cherkassky, Yunqian Ma

Electrical and Computer Engineering

Research output: Contribution to journal › Article › peer-review

53 Scopus citations

Abstract

This paper presents a new learning formulation for multiple model estimation (MME). Under this formulation, training data samples are generated by several (unknown) statistical models. Hence, most existing learning methods (for classification or regression) based on a single model formulation are no longer applicable. We describe a general framework for MME. Then we introduce a constructive support vector machine (SVM)-based methodology for multiple regression estimation. Several empirical comparisons using synthetic and real-life data sets are presented to illustrate the proposed approach for multiple model regression formulation.

Original language	English (US)
Pages (from-to)	785-798
Number of pages	14
Journal	IEEE Transactions on Neural Networks
Volume	16
Issue number	4
DOIs	https://doi.org/10.1109/TNN.2005.849836
State	Published - Jul 2005

Bibliographical note

Funding Information:
Manuscript received July 24, 2002; revised August 27, 2004. This work was supported in part by the National Science Foundation under Grant ECS-0099906 and in part by Fair Isaac Corporation.

Keywords

Learning formulation
Multiple model estimation (MME)
Regression
Robust estimation
Support vector machines (SVMs)

Publisher link

10.1109/TNN.2005.849836

Find It

Find It at U of M Libraries

Cite this

@article{0b8341e8fc044213a9c03cbd8d8e10a1,

title = "Multiple model regression estimation",

abstract = "This paper presents a new learning formulation for multiple model estimation (MME). Under this formulation, training data samples are generated by several (unknown) statistical models. Hence, most existing learning methods (for classification or regression) based on a single model formulation are no longer applicable. We describe a general framework for MME. Then we introduce a constructive support vector machine (SVM)-based methodology for multiple regression estimation. Several empirical comparisons using synthetic and real-life data sets are presented to illustrate the proposed approach for multiple model regression formulation.",

keywords = "Learning formulation, Multiple model estimation (MME), Regression, Robust estimation, Support vector machines (SVMs)",

author = "Vladimir Cherkassky and Yunqian Ma",

note = "Funding Information: Manuscript received July 24, 2002; revised August 27, 2004. This work was supported in part by the National Science Foundation under Grant ECS-0099906 and in part by Fair Isaac Corporation.",

year = "2005",

month = jul,

doi = "10.1109/TNN.2005.849836",

language = "English (US)",

volume = "16",

pages = "785--798",

journal = "IEEE Transactions on Neural Networks",

issn = "1045-9227",

publisher = "IEEE Computational Intelligence Society",

number = "4",

}

TY - JOUR

T1 - Multiple model regression estimation

AU - Cherkassky, Vladimir

AU - Ma, Yunqian

N1 - Funding Information: Manuscript received July 24, 2002; revised August 27, 2004. This work was supported in part by the National Science Foundation under Grant ECS-0099906 and in part by Fair Isaac Corporation.

PY - 2005/7

Y1 - 2005/7

N2 - This paper presents a new learning formulation for multiple model estimation (MME). Under this formulation, training data samples are generated by several (unknown) statistical models. Hence, most existing learning methods (for classification or regression) based on a single model formulation are no longer applicable. We describe a general framework for MME. Then we introduce a constructive support vector machine (SVM)-based methodology for multiple regression estimation. Several empirical comparisons using synthetic and real-life data sets are presented to illustrate the proposed approach for multiple model regression formulation.

AB - This paper presents a new learning formulation for multiple model estimation (MME). Under this formulation, training data samples are generated by several (unknown) statistical models. Hence, most existing learning methods (for classification or regression) based on a single model formulation are no longer applicable. We describe a general framework for MME. Then we introduce a constructive support vector machine (SVM)-based methodology for multiple regression estimation. Several empirical comparisons using synthetic and real-life data sets are presented to illustrate the proposed approach for multiple model regression formulation.

KW - Learning formulation

KW - Multiple model estimation (MME)

KW - Regression

KW - Robust estimation

KW - Support vector machines (SVMs)

UR - http://www.scopus.com/inward/record.url?scp=23044466683&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=23044466683&partnerID=8YFLogxK

U2 - 10.1109/TNN.2005.849836

DO - 10.1109/TNN.2005.849836

M3 - Article

C2 - 16121721

AN - SCOPUS:23044466683

SN - 1045-9227

VL - 16

SP - 785

EP - 798

JO - IEEE Transactions on Neural Networks

JF - IEEE Transactions on Neural Networks

IS - 4

ER -

Multiple model regression estimation

Abstract

Bibliographical note

Keywords

Publisher link

Find It

Other files and links

Fingerprint

Cite this