Propagating Distributions on a Hypergraph by Dual Information Regularization

Abstract

In the information regularization framework by Corduneanu and Jaakkola (2005), the distributions of labels are propagated on a hypergraph for semi-supervised learning. The learning is efficiently done by a Blahut-Arimoto-like two step algorithm, but, unfortunately, one of the steps cannot be solved in a closed form. In this paper, we propose a dual version of information regularization, which is considered as more natural in terms of information geometry. Our learning algorithm has two steps, each of which can be solved in a closed form. Also it can be naturally applied to exponential family distributions such as Gaussians. In experiments, our algorithm is applied to protein classification based on a metabolic network and known functional categories.

Cite

Text

Tsuda. "Propagating Distributions on a Hypergraph by Dual Information Regularization." International Conference on Machine Learning, 2005. doi:10.1145/1102351.1102467

Markdown

[Tsuda. "Propagating Distributions on a Hypergraph by Dual Information Regularization." International Conference on Machine Learning, 2005.](https://mlanthology.org/icml/2005/tsuda2005icml-propagating/) doi:10.1145/1102351.1102467

BibTeX

@inproceedings{tsuda2005icml-propagating,
  title     = {{Propagating Distributions on a Hypergraph by Dual Information Regularization}},
  author    = {Tsuda, Koji},
  booktitle = {International Conference on Machine Learning},
  year      = {2005},
  pages     = {920-927},
  doi       = {10.1145/1102351.1102467},
  url       = {https://mlanthology.org/icml/2005/tsuda2005icml-propagating/}
}