Combination of supervised and semi-supervised regression models for improved unbiased estimation

Jeronimo Arenas-Garía; Carlos Moriana-Varo; Jan Larsen

Combination of supervised and semi-supervised regression models for improved unbiased estimation

Jeronimo Arenas-Garía, Carlos Moriana-Varo, Jan Larsen

Universidad Carlos III de Madrid

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

130 Downloads (Pure)

Abstract

In this paper we investigate the steady-state performance of semisupervised regression models adjusted using a modified RLS-like algorithm, identifying the situations where the new algorithm is expected to outperform standard RLS. By using an adaptive combination of the supervised and semisupervised methods, the resulting adaptive filter is guaranteed to perform at least as well as the best contributing filter, therefore achieving universal performance. The analysis and behavior of the methods is illustrated through a set of examples in a plant identification setup, analyzing both steady-state and convergence situations.

Original language	English
Title of host publication	Proceedings of the Seventh International Symposium on Wireless Communication Systems (ISWCS2010)
Publisher	IEEE Press
Publication date	2010
Publication status	Published - 2010
Event	Seventh International Symposium on Wireless Communication Systems : ISWCS2010 - Duration: 1 Jan 2010 → …

Conference

Conference	Seventh International Symposium on Wireless Communication Systems : ISWCS2010
Period	01/01/2010 → …

Keywords

semi-supervised learning

Access to Document

6729anavSubmitted manuscript, 202 KB

http://ieeexplore.ieee.org

OpenUrl availability

Full text

Cite this

@inproceedings{4ff40fe62c2c448cb05d87b17d91e91e,

title = "Combination of supervised and semi-supervised regression models for improved unbiased estimation",

abstract = "In this paper we investigate the steady-state performance of semisupervised regression models adjusted using a modified RLS-like algorithm, identifying the situations where the new algorithm is expected to outperform standard RLS. By using an adaptive combination of the supervised and semisupervised methods, the resulting adaptive filter is guaranteed to perform at least as well as the best contributing filter, therefore achieving universal performance. The analysis and behavior of the methods is illustrated through a set of examples in a plant identification setup, analyzing both steady-state and convergence situations.",

keywords = "semi-supervised learning",

author = "Jeronimo Arenas-Gar{\'i}a and Carlos Moriana-Varo and Jan Larsen",

year = "2010",

language = "English",

booktitle = "Proceedings of the Seventh International Symposium on Wireless Communication Systems (ISWCS2010)",

publisher = "IEEE Press",

note = "Seventh International Symposium on Wireless Communication Systems : ISWCS2010 ; Conference date: 01-01-2010",

}

Arenas-Garía, J, Moriana-Varo, C & Larsen, J 2010, Combination of supervised and semi-supervised regression models for improved unbiased estimation. in Proceedings of the Seventh International Symposium on Wireless Communication Systems (ISWCS2010). IEEE Press, Seventh International Symposium on Wireless Communication Systems : ISWCS2010, 01/01/2010. <http://ieeexplore.ieee.org>

Combination of supervised and semi-supervised regression models for improved unbiased estimation. / Arenas-Garía, Jeronimo; Moriana-Varo, Carlos; Larsen, Jan.
Proceedings of the Seventh International Symposium on Wireless Communication Systems (ISWCS2010). IEEE Press, 2010.

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

TY - GEN

T1 - Combination of supervised and semi-supervised regression models for improved unbiased estimation

AU - Arenas-Garía, Jeronimo

AU - Moriana-Varo, Carlos

AU - Larsen, Jan

PY - 2010

Y1 - 2010

N2 - In this paper we investigate the steady-state performance of semisupervised regression models adjusted using a modified RLS-like algorithm, identifying the situations where the new algorithm is expected to outperform standard RLS. By using an adaptive combination of the supervised and semisupervised methods, the resulting adaptive filter is guaranteed to perform at least as well as the best contributing filter, therefore achieving universal performance. The analysis and behavior of the methods is illustrated through a set of examples in a plant identification setup, analyzing both steady-state and convergence situations.

AB - In this paper we investigate the steady-state performance of semisupervised regression models adjusted using a modified RLS-like algorithm, identifying the situations where the new algorithm is expected to outperform standard RLS. By using an adaptive combination of the supervised and semisupervised methods, the resulting adaptive filter is guaranteed to perform at least as well as the best contributing filter, therefore achieving universal performance. The analysis and behavior of the methods is illustrated through a set of examples in a plant identification setup, analyzing both steady-state and convergence situations.

KW - semi-supervised learning

M3 - Article in proceedings

BT - Proceedings of the Seventh International Symposium on Wireless Communication Systems (ISWCS2010)

PB - IEEE Press

T2 - Seventh International Symposium on Wireless Communication Systems : ISWCS2010

Y2 - 1 January 2010

ER -

Combination of supervised and semi-supervised regression models for improved unbiased estimation

Abstract

Conference

Keywords

Access to Document

OpenUrl availability

Fingerprint

Cite this