Gaussian process for dimensionality reduction in transfer learning

Bin Tong, Junbin Gao, Nguyen Huy Thach, Einoshin Suzuki

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

5 Citations (Scopus)

Abstract

Dimensionality reduction has been considered as one of the most significant tools for data analysis. In general, supervised information is helpful for dimensionality reduction. However, in typical real applications, supervised information in multiple source tasks may be available, while the data of the target task are unlabeled. An interesting problem of how to guide the dimensionality reduction for the unlabeled target data by exploiting useful knowledge, such as label information, from multiple source tasks arises in such a scenario. In this paper, we propose a new method for dimensionality reduction in the transfer learning setting. Unlike traditional paradigms where the useful knowledge from multiple source tasks is transferred through distance metric, our proposal firstly converts the dimensionality reduction problem into integral regression problems in parallel. Gaussian process is then employed to learn the underlying relationship between the original data and the reduced data. Such a relationship can be appropriately transferred to the target task by exploiting the prediction ability of the Gaussian process model and inventing different kinds of regularizers. Extensive experiments on both synthetic and real data sets show the effectiveness of our method.

Original languageEnglish
Title of host publicationProceedings of the 11th SIAM International Conference on Data Mining, SDM 2011
Pages783-794
Number of pages12
Publication statusPublished - Dec 1 2011
Event11th SIAM International Conference on Data Mining, SDM 2011 - Mesa, AZ, United States
Duration: Apr 28 2011Apr 30 2011

Publication series

NameProceedings of the 11th SIAM International Conference on Data Mining, SDM 2011

Other

Other11th SIAM International Conference on Data Mining, SDM 2011
CountryUnited States
CityMesa, AZ
Period4/28/114/30/11

Fingerprint

Labels
Experiments

All Science Journal Classification (ASJC) codes

  • Software

Cite this

Tong, B., Gao, J., Thach, N. H., & Suzuki, E. (2011). Gaussian process for dimensionality reduction in transfer learning. In Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011 (pp. 783-794). (Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011).

Gaussian process for dimensionality reduction in transfer learning. / Tong, Bin; Gao, Junbin; Thach, Nguyen Huy; Suzuki, Einoshin.

Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011. 2011. p. 783-794 (Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011).

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

Tong, B, Gao, J, Thach, NH & Suzuki, E 2011, Gaussian process for dimensionality reduction in transfer learning. in Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011. Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011, pp. 783-794, 11th SIAM International Conference on Data Mining, SDM 2011, Mesa, AZ, United States, 4/28/11.
Tong B, Gao J, Thach NH, Suzuki E. Gaussian process for dimensionality reduction in transfer learning. In Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011. 2011. p. 783-794. (Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011).
Tong, Bin ; Gao, Junbin ; Thach, Nguyen Huy ; Suzuki, Einoshin. / Gaussian process for dimensionality reduction in transfer learning. Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011. 2011. pp. 783-794 (Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011).
@inproceedings{573e4a3c9a48464abb9e4e5946881f39,
title = "Gaussian process for dimensionality reduction in transfer learning",
abstract = "Dimensionality reduction has been considered as one of the most significant tools for data analysis. In general, supervised information is helpful for dimensionality reduction. However, in typical real applications, supervised information in multiple source tasks may be available, while the data of the target task are unlabeled. An interesting problem of how to guide the dimensionality reduction for the unlabeled target data by exploiting useful knowledge, such as label information, from multiple source tasks arises in such a scenario. In this paper, we propose a new method for dimensionality reduction in the transfer learning setting. Unlike traditional paradigms where the useful knowledge from multiple source tasks is transferred through distance metric, our proposal firstly converts the dimensionality reduction problem into integral regression problems in parallel. Gaussian process is then employed to learn the underlying relationship between the original data and the reduced data. Such a relationship can be appropriately transferred to the target task by exploiting the prediction ability of the Gaussian process model and inventing different kinds of regularizers. Extensive experiments on both synthetic and real data sets show the effectiveness of our method.",
author = "Bin Tong and Junbin Gao and Thach, {Nguyen Huy} and Einoshin Suzuki",
year = "2011",
month = "12",
day = "1",
language = "English",
isbn = "9780898719925",
series = "Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011",
pages = "783--794",
booktitle = "Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011",

}

TY - GEN

T1 - Gaussian process for dimensionality reduction in transfer learning

AU - Tong, Bin

AU - Gao, Junbin

AU - Thach, Nguyen Huy

AU - Suzuki, Einoshin

PY - 2011/12/1

Y1 - 2011/12/1

N2 - Dimensionality reduction has been considered as one of the most significant tools for data analysis. In general, supervised information is helpful for dimensionality reduction. However, in typical real applications, supervised information in multiple source tasks may be available, while the data of the target task are unlabeled. An interesting problem of how to guide the dimensionality reduction for the unlabeled target data by exploiting useful knowledge, such as label information, from multiple source tasks arises in such a scenario. In this paper, we propose a new method for dimensionality reduction in the transfer learning setting. Unlike traditional paradigms where the useful knowledge from multiple source tasks is transferred through distance metric, our proposal firstly converts the dimensionality reduction problem into integral regression problems in parallel. Gaussian process is then employed to learn the underlying relationship between the original data and the reduced data. Such a relationship can be appropriately transferred to the target task by exploiting the prediction ability of the Gaussian process model and inventing different kinds of regularizers. Extensive experiments on both synthetic and real data sets show the effectiveness of our method.

AB - Dimensionality reduction has been considered as one of the most significant tools for data analysis. In general, supervised information is helpful for dimensionality reduction. However, in typical real applications, supervised information in multiple source tasks may be available, while the data of the target task are unlabeled. An interesting problem of how to guide the dimensionality reduction for the unlabeled target data by exploiting useful knowledge, such as label information, from multiple source tasks arises in such a scenario. In this paper, we propose a new method for dimensionality reduction in the transfer learning setting. Unlike traditional paradigms where the useful knowledge from multiple source tasks is transferred through distance metric, our proposal firstly converts the dimensionality reduction problem into integral regression problems in parallel. Gaussian process is then employed to learn the underlying relationship between the original data and the reduced data. Such a relationship can be appropriately transferred to the target task by exploiting the prediction ability of the Gaussian process model and inventing different kinds of regularizers. Extensive experiments on both synthetic and real data sets show the effectiveness of our method.

UR - http://www.scopus.com/inward/record.url?scp=84866454082&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84866454082&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84866454082

SN - 9780898719925

T3 - Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011

SP - 783

EP - 794

BT - Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011

ER -