Semi-supervised Gaussian process regression and its feedback design

Xinlu Guo, Yoshiaki Yasumura, Kuniaki Uehara

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

1 Citation (Scopus)

Abstract

Semi-supervised learning has received considerable attention in the machine learning literature due to its potential in reducing the need for expensive labeled data. The majority of the proposed algorithms, however, have been applied to the classification task. In this paper we present a graph-based semi-supervised algorithm for solving regression problem. Our method incorporates an adjacent graph, which is built on labeled and unlabeled data, with the standard Gaussian process (GP) prior to infer the new training and predicting distribution for semi-supervised GP regression (GPr). Additionally, in semi-supervised regression, the prediction of unlabeled data could contain some valuable information. For example, it can be seen as labeled data paired with the unlabeled data, and under some metrics, they can help to construct more accurate model. Therefore, we also describe a feedback algorithm, which can choose the useful prediction of unlabeled data for feedback to re-train the model iteratively. Experimental results show that our work achieves comparable performance to standard GPr.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages353-366
Number of pages14
Volume7713 LNAI
DOIs
Publication statusPublished - 2012
Event8th International Conference on Advanced Data Mining and Applications, ADMA 2012 - Nanjing
Duration: 2012 Dec 152012 Dec 18

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume7713 LNAI
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other8th International Conference on Advanced Data Mining and Applications, ADMA 2012
CityNanjing
Period12/12/1512/12/18

Fingerprint

Feedback
Supervised learning
Learning systems

Keywords

  • Feedback
  • Gaussian process
  • Graph laplacian
  • Regression
  • Semi-supervised learning

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Guo, X., Yasumura, Y., & Uehara, K. (2012). Semi-supervised Gaussian process regression and its feedback design. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7713 LNAI, pp. 353-366). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7713 LNAI). https://doi.org/10.1007/978-3-642-35527-1_30

Semi-supervised Gaussian process regression and its feedback design. / Guo, Xinlu; Yasumura, Yoshiaki; Uehara, Kuniaki.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7713 LNAI 2012. p. 353-366 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7713 LNAI).

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

Guo, X, Yasumura, Y & Uehara, K 2012, Semi-supervised Gaussian process regression and its feedback design. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 7713 LNAI, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7713 LNAI, pp. 353-366, 8th International Conference on Advanced Data Mining and Applications, ADMA 2012, Nanjing, 12/12/15. https://doi.org/10.1007/978-3-642-35527-1_30
Guo X, Yasumura Y, Uehara K. Semi-supervised Gaussian process regression and its feedback design. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7713 LNAI. 2012. p. 353-366. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-35527-1_30
Guo, Xinlu ; Yasumura, Yoshiaki ; Uehara, Kuniaki. / Semi-supervised Gaussian process regression and its feedback design. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7713 LNAI 2012. pp. 353-366 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{183e93da0034426aa2436c865d86d500,
title = "Semi-supervised Gaussian process regression and its feedback design",
abstract = "Semi-supervised learning has received considerable attention in the machine learning literature due to its potential in reducing the need for expensive labeled data. The majority of the proposed algorithms, however, have been applied to the classification task. In this paper we present a graph-based semi-supervised algorithm for solving regression problem. Our method incorporates an adjacent graph, which is built on labeled and unlabeled data, with the standard Gaussian process (GP) prior to infer the new training and predicting distribution for semi-supervised GP regression (GPr). Additionally, in semi-supervised regression, the prediction of unlabeled data could contain some valuable information. For example, it can be seen as labeled data paired with the unlabeled data, and under some metrics, they can help to construct more accurate model. Therefore, we also describe a feedback algorithm, which can choose the useful prediction of unlabeled data for feedback to re-train the model iteratively. Experimental results show that our work achieves comparable performance to standard GPr.",
keywords = "Feedback, Gaussian process, Graph laplacian, Regression, Semi-supervised learning",
author = "Xinlu Guo and Yoshiaki Yasumura and Kuniaki Uehara",
year = "2012",
doi = "10.1007/978-3-642-35527-1_30",
language = "English",
isbn = "9783642355264",
volume = "7713 LNAI",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "353--366",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Semi-supervised Gaussian process regression and its feedback design

AU - Guo, Xinlu

AU - Yasumura, Yoshiaki

AU - Uehara, Kuniaki

PY - 2012

Y1 - 2012

N2 - Semi-supervised learning has received considerable attention in the machine learning literature due to its potential in reducing the need for expensive labeled data. The majority of the proposed algorithms, however, have been applied to the classification task. In this paper we present a graph-based semi-supervised algorithm for solving regression problem. Our method incorporates an adjacent graph, which is built on labeled and unlabeled data, with the standard Gaussian process (GP) prior to infer the new training and predicting distribution for semi-supervised GP regression (GPr). Additionally, in semi-supervised regression, the prediction of unlabeled data could contain some valuable information. For example, it can be seen as labeled data paired with the unlabeled data, and under some metrics, they can help to construct more accurate model. Therefore, we also describe a feedback algorithm, which can choose the useful prediction of unlabeled data for feedback to re-train the model iteratively. Experimental results show that our work achieves comparable performance to standard GPr.

AB - Semi-supervised learning has received considerable attention in the machine learning literature due to its potential in reducing the need for expensive labeled data. The majority of the proposed algorithms, however, have been applied to the classification task. In this paper we present a graph-based semi-supervised algorithm for solving regression problem. Our method incorporates an adjacent graph, which is built on labeled and unlabeled data, with the standard Gaussian process (GP) prior to infer the new training and predicting distribution for semi-supervised GP regression (GPr). Additionally, in semi-supervised regression, the prediction of unlabeled data could contain some valuable information. For example, it can be seen as labeled data paired with the unlabeled data, and under some metrics, they can help to construct more accurate model. Therefore, we also describe a feedback algorithm, which can choose the useful prediction of unlabeled data for feedback to re-train the model iteratively. Experimental results show that our work achieves comparable performance to standard GPr.

KW - Feedback

KW - Gaussian process

KW - Graph laplacian

KW - Regression

KW - Semi-supervised learning

UR - http://www.scopus.com/inward/record.url?scp=84872690376&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84872690376&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-35527-1_30

DO - 10.1007/978-3-642-35527-1_30

M3 - Conference contribution

AN - SCOPUS:84872690376

SN - 9783642355264

VL - 7713 LNAI

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 353

EP - 366

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

ER -