DP-space: Bayesian Nonparametric Subspace Clustering with Small-variance Asymptotics

Yining Wang, Jun Zhu
Proceedings of the 32nd International Conference on Machine Learning, PMLR 37:862-870, 2015.

Abstract

Subspace clustering separates data points approximately lying on union of affine subspaces into several clusters. This paper presents a novel nonparametric Bayesian subspace clustering model that infers both the number of subspaces and the dimension of each subspace from the observed data. Though the posterior inference is hard, our model leads to a very efficient deterministic algorithm, DP-space, which retains the nonparametric ability under a small-variance asymptotic analysis. DP-space monotonically minimizes an intuitive objective with an explicit tradeoff between data fitness and model complexity. Experimental results demonstrate that DP-space outperforms various competitors in terms of clustering accuracy and at the same time it is highly efficient.

Cite this Paper

BibTeX
@InProceedings{pmlr-v37-wanga15, title = {DP-space: Bayesian Nonparametric Subspace Clustering with Small-variance Asymptotics}, author = {Wang, Yining and Zhu, Jun}, booktitle = {Proceedings of the 32nd International Conference on Machine Learning}, pages = {862--870}, year = {2015}, editor = {Bach, Francis and Blei, David}, volume = {37}, series = {Proceedings of Machine Learning Research}, address = {Lille, France}, month = {07--09 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v37/wanga15.pdf}, url = {https://proceedings.mlr.press/v37/wanga15.html}, abstract = {Subspace clustering separates data points approximately lying on union of affine subspaces into several clusters. This paper presents a novel nonparametric Bayesian subspace clustering model that infers both the number of subspaces and the dimension of each subspace from the observed data. Though the posterior inference is hard, our model leads to a very efficient deterministic algorithm, DP-space, which retains the nonparametric ability under a small-variance asymptotic analysis. DP-space monotonically minimizes an intuitive objective with an explicit tradeoff between data fitness and model complexity. Experimental results demonstrate that DP-space outperforms various competitors in terms of clustering accuracy and at the same time it is highly efficient.} }
Endnote
%0 Conference Paper %T DP-space: Bayesian Nonparametric Subspace Clustering with Small-variance Asymptotics %A Yining Wang %A Jun Zhu %B Proceedings of the 32nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2015 %E Francis Bach %E David Blei %F pmlr-v37-wanga15 %I PMLR %P 862--870 %U https://proceedings.mlr.press/v37/wanga15.html %V 37 %X Subspace clustering separates data points approximately lying on union of affine subspaces into several clusters. This paper presents a novel nonparametric Bayesian subspace clustering model that infers both the number of subspaces and the dimension of each subspace from the observed data. Though the posterior inference is hard, our model leads to a very efficient deterministic algorithm, DP-space, which retains the nonparametric ability under a small-variance asymptotic analysis. DP-space monotonically minimizes an intuitive objective with an explicit tradeoff between data fitness and model complexity. Experimental results demonstrate that DP-space outperforms various competitors in terms of clustering accuracy and at the same time it is highly efficient.
RIS
TY - CPAPER TI - DP-space: Bayesian Nonparametric Subspace Clustering with Small-variance Asymptotics AU - Yining Wang AU - Jun Zhu BT - Proceedings of the 32nd International Conference on Machine Learning DA - 2015/06/01 ED - Francis Bach ED - David Blei ID - pmlr-v37-wanga15 PB - PMLR DP - Proceedings of Machine Learning Research VL - 37 SP - 862 EP - 870 L1 - http://proceedings.mlr.press/v37/wanga15.pdf UR - https://proceedings.mlr.press/v37/wanga15.html AB - Subspace clustering separates data points approximately lying on union of affine subspaces into several clusters. This paper presents a novel nonparametric Bayesian subspace clustering model that infers both the number of subspaces and the dimension of each subspace from the observed data. Though the posterior inference is hard, our model leads to a very efficient deterministic algorithm, DP-space, which retains the nonparametric ability under a small-variance asymptotic analysis. DP-space monotonically minimizes an intuitive objective with an explicit tradeoff between data fitness and model complexity. Experimental results demonstrate that DP-space outperforms various competitors in terms of clustering accuracy and at the same time it is highly efficient. ER -
APA
Wang, Y. & Zhu, J.. (2015). DP-space: Bayesian Nonparametric Subspace Clustering with Small-variance Asymptotics. Proceedings of the 32nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 37:862-870 Available from https://proceedings.mlr.press/v37/wanga15.html.

Related Material