Semi-Supervised Mean Fields

Fei Wang, Shijun Wang, Changshui Zhang, Ole Winther
; Proceedings of the Eleventh International Conference on Artificial Intelligence and Statistics, PMLR 2:596-603, 2007.

Abstract

A novel semi-supervised learning approach based on statistical physics is proposed in this paper. We treat each data point as an Ising spin and the interaction between pairwise spins is captured by the similarity between the pairwise points. The labels of the data points are treated as the directions of the corresponding spins. In semi-supervised setting, some of the spins have fixed directions (which corresponds to the labeled data), and our task is to determine the directions of other spins. An approach based on the Mean Field theory is proposed to achieve this goal. Finally the experimental results on both toy and real world data sets are provided to show the effectiveness of our method.

Cite this Paper


BibTeX
@InProceedings{pmlr-v2-wang07c, title = {Semi-Supervised Mean Fields}, author = {Fei Wang and Shijun Wang and Changshui Zhang and Ole Winther}, booktitle = {Proceedings of the Eleventh International Conference on Artificial Intelligence and Statistics}, pages = {596--603}, year = {2007}, editor = {Marina Meila and Xiaotong Shen}, volume = {2}, series = {Proceedings of Machine Learning Research}, address = {San Juan, Puerto Rico}, month = {21--24 Mar}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v2/wang07c/wang07c.pdf}, url = {http://proceedings.mlr.press/v2/wang07c.html}, abstract = {A novel semi-supervised learning approach based on statistical physics is proposed in this paper. We treat each data point as an Ising spin and the interaction between pairwise spins is captured by the similarity between the pairwise points. The labels of the data points are treated as the directions of the corresponding spins. In semi-supervised setting, some of the spins have fixed directions (which corresponds to the labeled data), and our task is to determine the directions of other spins. An approach based on the Mean Field theory is proposed to achieve this goal. Finally the experimental results on both toy and real world data sets are provided to show the effectiveness of our method.} }
Endnote
%0 Conference Paper %T Semi-Supervised Mean Fields %A Fei Wang %A Shijun Wang %A Changshui Zhang %A Ole Winther %B Proceedings of the Eleventh International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2007 %E Marina Meila %E Xiaotong Shen %F pmlr-v2-wang07c %I PMLR %J Proceedings of Machine Learning Research %P 596--603 %U http://proceedings.mlr.press %V 2 %W PMLR %X A novel semi-supervised learning approach based on statistical physics is proposed in this paper. We treat each data point as an Ising spin and the interaction between pairwise spins is captured by the similarity between the pairwise points. The labels of the data points are treated as the directions of the corresponding spins. In semi-supervised setting, some of the spins have fixed directions (which corresponds to the labeled data), and our task is to determine the directions of other spins. An approach based on the Mean Field theory is proposed to achieve this goal. Finally the experimental results on both toy and real world data sets are provided to show the effectiveness of our method.
RIS
TY - CPAPER TI - Semi-Supervised Mean Fields AU - Fei Wang AU - Shijun Wang AU - Changshui Zhang AU - Ole Winther BT - Proceedings of the Eleventh International Conference on Artificial Intelligence and Statistics PY - 2007/03/11 DA - 2007/03/11 ED - Marina Meila ED - Xiaotong Shen ID - pmlr-v2-wang07c PB - PMLR SP - 596 DP - PMLR EP - 603 L1 - http://proceedings.mlr.press/v2/wang07c/wang07c.pdf UR - http://proceedings.mlr.press/v2/wang07c.html AB - A novel semi-supervised learning approach based on statistical physics is proposed in this paper. We treat each data point as an Ising spin and the interaction between pairwise spins is captured by the similarity between the pairwise points. The labels of the data points are treated as the directions of the corresponding spins. In semi-supervised setting, some of the spins have fixed directions (which corresponds to the labeled data), and our task is to determine the directions of other spins. An approach based on the Mean Field theory is proposed to achieve this goal. Finally the experimental results on both toy and real world data sets are provided to show the effectiveness of our method. ER -
APA
Wang, F., Wang, S., Zhang, C. & Winther, O.. (2007). Semi-Supervised Mean Fields. Proceedings of the Eleventh International Conference on Artificial Intelligence and Statistics, in PMLR 2:596-603

Related Material