Home > Research > Publications & Outputs > Subspace Clustering with Active Learning

Electronic data

  • BigData2019

    Accepted author manuscript, 352 KB, PDF document

Links

Text available via DOI:

View graph of relations

Subspace Clustering with Active Learning

Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSNConference contribution/Paper

Published

Standard

Subspace Clustering with Active Learning. / Peng, Hankui; Pavlidis, Nicos.

Proceedings of 2019 IEEE International Conference on Big Data. IEEE, 2020. 9006361.

Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSNConference contribution/Paper

Harvard

Peng, H & Pavlidis, N 2020, Subspace Clustering with Active Learning. in Proceedings of 2019 IEEE International Conference on Big Data., 9006361, IEEE, IEEE International Conference on Big Data, Los Angeles, United States, 9/12/19. https://doi.org/10.1109/BigData47090.2019.9006361

APA

Peng, H., & Pavlidis, N. (2020). Subspace Clustering with Active Learning. In Proceedings of 2019 IEEE International Conference on Big Data [9006361] IEEE. https://doi.org/10.1109/BigData47090.2019.9006361

Vancouver

Peng H, Pavlidis N. Subspace Clustering with Active Learning. In Proceedings of 2019 IEEE International Conference on Big Data. IEEE. 2020. 9006361 https://doi.org/10.1109/BigData47090.2019.9006361

Author

Peng, Hankui ; Pavlidis, Nicos. / Subspace Clustering with Active Learning. Proceedings of 2019 IEEE International Conference on Big Data. IEEE, 2020.

Bibtex

@inproceedings{9ee0cb275d7d4814b167212cfa16cbc4,
title = "Subspace Clustering with Active Learning",
abstract = "Subspace clustering is a growing field of unsupervised learning that has gained much popularity in the computer vision community. Applications can be found in areas such as motion segmentation and face clustering. It assumes that data originate from a union of subspaces, and clusters the data depending on the corresponding subspace. In practice, it is reasonable to assume that a limited amount of labels can be obtained, potentially at a cost. Therefore, algorithms that can effectively and efficiently incorporate this information to improve the clustering model are desirable. In this paper, we propose an active learning framework for subspace clustering that sequentially queries informative points and updates the subspace model. The query stage of the proposed frameworkrelies on results from the perturbation theory of principal component analysis, to identify influential and potentially misclassified points. A constrained subspace clustering algorithm is proposed that monotonically decreases the objective function subject to the constraints imposed by the labelled data. We show that our proposed framework is suitable for subspace clusteringalgorithms including iterative methods and spectral methods. Experiments on synthetic data sets, motion segmentation data sets, and Yale Faces data sets demonstrate the advantage of our proposed active strategy over state-of-the-art.",
author = "Hankui Peng and Nicos Pavlidis",
year = "2020",
month = feb,
day = "24",
doi = "10.1109/BigData47090.2019.9006361",
language = "English",
booktitle = "Proceedings of 2019 IEEE International Conference on Big Data",
publisher = "IEEE",
note = "IEEE International Conference on Big Data, IEEE BigData 2019 ; Conference date: 09-12-2019 Through 12-12-2019",
url = "http://bigdataieee.org/BigData2019/index.html",

}

RIS

TY - GEN

T1 - Subspace Clustering with Active Learning

AU - Peng, Hankui

AU - Pavlidis, Nicos

PY - 2020/2/24

Y1 - 2020/2/24

N2 - Subspace clustering is a growing field of unsupervised learning that has gained much popularity in the computer vision community. Applications can be found in areas such as motion segmentation and face clustering. It assumes that data originate from a union of subspaces, and clusters the data depending on the corresponding subspace. In practice, it is reasonable to assume that a limited amount of labels can be obtained, potentially at a cost. Therefore, algorithms that can effectively and efficiently incorporate this information to improve the clustering model are desirable. In this paper, we propose an active learning framework for subspace clustering that sequentially queries informative points and updates the subspace model. The query stage of the proposed frameworkrelies on results from the perturbation theory of principal component analysis, to identify influential and potentially misclassified points. A constrained subspace clustering algorithm is proposed that monotonically decreases the objective function subject to the constraints imposed by the labelled data. We show that our proposed framework is suitable for subspace clusteringalgorithms including iterative methods and spectral methods. Experiments on synthetic data sets, motion segmentation data sets, and Yale Faces data sets demonstrate the advantage of our proposed active strategy over state-of-the-art.

AB - Subspace clustering is a growing field of unsupervised learning that has gained much popularity in the computer vision community. Applications can be found in areas such as motion segmentation and face clustering. It assumes that data originate from a union of subspaces, and clusters the data depending on the corresponding subspace. In practice, it is reasonable to assume that a limited amount of labels can be obtained, potentially at a cost. Therefore, algorithms that can effectively and efficiently incorporate this information to improve the clustering model are desirable. In this paper, we propose an active learning framework for subspace clustering that sequentially queries informative points and updates the subspace model. The query stage of the proposed frameworkrelies on results from the perturbation theory of principal component analysis, to identify influential and potentially misclassified points. A constrained subspace clustering algorithm is proposed that monotonically decreases the objective function subject to the constraints imposed by the labelled data. We show that our proposed framework is suitable for subspace clusteringalgorithms including iterative methods and spectral methods. Experiments on synthetic data sets, motion segmentation data sets, and Yale Faces data sets demonstrate the advantage of our proposed active strategy over state-of-the-art.

U2 - 10.1109/BigData47090.2019.9006361

DO - 10.1109/BigData47090.2019.9006361

M3 - Conference contribution/Paper

BT - Proceedings of 2019 IEEE International Conference on Big Data

PB - IEEE

T2 - IEEE International Conference on Big Data

Y2 - 9 December 2019 through 12 December 2019

ER -