Partial least squares regression for graph mining

Hiroto Saigo, Nicole Krämer, Koji Tsuda

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

51 Citations (Scopus)

Abstract

Attributed graphs are increasingly more common in many application domains such as chemistry, biology and text processing. A central issue in graph mining is how to collect informative subgraph patterns for a given learning task. We propose an iterative mining method based on partial least squares regression (PLS). To apply PLS to graph data, a sparse version of PLS is developed first and then it is combined with a weighted pattern mining algorithm. The mining algorithm is iteratively called with different weight vectors, creating one latent component per one mining call. Our method, graph PLS, is efficient and easy to implement, because the weight vector is updated with elementary matrix calculations. In experiments, our graph PLS algorithm showed competitive prediction accuracies in many chemical datasets and its efficiency was significantly superior to graph boosting (gBoost) and the naive method based on frequent graph mining.

Original languageEnglish
Title of host publicationKDD 2008 - Proceedings of the 14th ACMKDD International Conference on Knowledge Discovery and Data Mining
Pages578-586
Number of pages9
DOIs
Publication statusPublished - Dec 1 2008
Externally publishedYes
Event14th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2008 - Las Vegas, NV, United States
Duration: Aug 24 2008Aug 27 2008

Publication series

NameProceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining

Other

Other14th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2008
CountryUnited States
CityLas Vegas, NV
Period8/24/088/27/08

Fingerprint

Text processing
Experiments

All Science Journal Classification (ASJC) codes

  • Software
  • Information Systems

Cite this

Saigo, H., Krämer, N., & Tsuda, K. (2008). Partial least squares regression for graph mining. In KDD 2008 - Proceedings of the 14th ACMKDD International Conference on Knowledge Discovery and Data Mining (pp. 578-586). (Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining). https://doi.org/10.1145/1401890.1401961

Partial least squares regression for graph mining. / Saigo, Hiroto; Krämer, Nicole; Tsuda, Koji.

KDD 2008 - Proceedings of the 14th ACMKDD International Conference on Knowledge Discovery and Data Mining. 2008. p. 578-586 (Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining).

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

Saigo, H, Krämer, N & Tsuda, K 2008, Partial least squares regression for graph mining. in KDD 2008 - Proceedings of the 14th ACMKDD International Conference on Knowledge Discovery and Data Mining. Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 578-586, 14th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2008, Las Vegas, NV, United States, 8/24/08. https://doi.org/10.1145/1401890.1401961
Saigo H, Krämer N, Tsuda K. Partial least squares regression for graph mining. In KDD 2008 - Proceedings of the 14th ACMKDD International Conference on Knowledge Discovery and Data Mining. 2008. p. 578-586. (Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining). https://doi.org/10.1145/1401890.1401961
Saigo, Hiroto ; Krämer, Nicole ; Tsuda, Koji. / Partial least squares regression for graph mining. KDD 2008 - Proceedings of the 14th ACMKDD International Conference on Knowledge Discovery and Data Mining. 2008. pp. 578-586 (Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining).
@inproceedings{9ac7109643a94b98ace6438ece01438e,
title = "Partial least squares regression for graph mining",
abstract = "Attributed graphs are increasingly more common in many application domains such as chemistry, biology and text processing. A central issue in graph mining is how to collect informative subgraph patterns for a given learning task. We propose an iterative mining method based on partial least squares regression (PLS). To apply PLS to graph data, a sparse version of PLS is developed first and then it is combined with a weighted pattern mining algorithm. The mining algorithm is iteratively called with different weight vectors, creating one latent component per one mining call. Our method, graph PLS, is efficient and easy to implement, because the weight vector is updated with elementary matrix calculations. In experiments, our graph PLS algorithm showed competitive prediction accuracies in many chemical datasets and its efficiency was significantly superior to graph boosting (gBoost) and the naive method based on frequent graph mining.",
author = "Hiroto Saigo and Nicole Kr{\"a}mer and Koji Tsuda",
year = "2008",
month = "12",
day = "1",
doi = "10.1145/1401890.1401961",
language = "English",
isbn = "9781605581934",
series = "Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining",
pages = "578--586",
booktitle = "KDD 2008 - Proceedings of the 14th ACMKDD International Conference on Knowledge Discovery and Data Mining",

}

TY - GEN

T1 - Partial least squares regression for graph mining

AU - Saigo, Hiroto

AU - Krämer, Nicole

AU - Tsuda, Koji

PY - 2008/12/1

Y1 - 2008/12/1

N2 - Attributed graphs are increasingly more common in many application domains such as chemistry, biology and text processing. A central issue in graph mining is how to collect informative subgraph patterns for a given learning task. We propose an iterative mining method based on partial least squares regression (PLS). To apply PLS to graph data, a sparse version of PLS is developed first and then it is combined with a weighted pattern mining algorithm. The mining algorithm is iteratively called with different weight vectors, creating one latent component per one mining call. Our method, graph PLS, is efficient and easy to implement, because the weight vector is updated with elementary matrix calculations. In experiments, our graph PLS algorithm showed competitive prediction accuracies in many chemical datasets and its efficiency was significantly superior to graph boosting (gBoost) and the naive method based on frequent graph mining.

AB - Attributed graphs are increasingly more common in many application domains such as chemistry, biology and text processing. A central issue in graph mining is how to collect informative subgraph patterns for a given learning task. We propose an iterative mining method based on partial least squares regression (PLS). To apply PLS to graph data, a sparse version of PLS is developed first and then it is combined with a weighted pattern mining algorithm. The mining algorithm is iteratively called with different weight vectors, creating one latent component per one mining call. Our method, graph PLS, is efficient and easy to implement, because the weight vector is updated with elementary matrix calculations. In experiments, our graph PLS algorithm showed competitive prediction accuracies in many chemical datasets and its efficiency was significantly superior to graph boosting (gBoost) and the naive method based on frequent graph mining.

UR - http://www.scopus.com/inward/record.url?scp=65449142148&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=65449142148&partnerID=8YFLogxK

U2 - 10.1145/1401890.1401961

DO - 10.1145/1401890.1401961

M3 - Conference contribution

AN - SCOPUS:65449142148

SN - 9781605581934

T3 - Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining

SP - 578

EP - 586

BT - KDD 2008 - Proceedings of the 14th ACMKDD International Conference on Knowledge Discovery and Data Mining

ER -