MedLDA: Maximum Margin Supervised Topic Models for Regression and Classification
Abstract
Supervised topic models utilize document's side information for discovering predictive low dimensional representations of documents; and existing models apply likelihood-based estimation. In this paper, we present a max-margin supervised topic model for both continuous and categorical response variables. Our approach, the maximum entropy discrimination latent Dirichlet allocation (MedLDA), utilizes the max-margin principle to train supervised topic models and estimate predictive topic representations that are arguably more suitable for prediction. We develop efficient variational methods for posterior inference and demonstrate qualitatively and quantitatively the advantages of MedLDA over likelihood-based topic models on movie review and 20 Newsgroups data sets.
Cite
Text
Zhu et al. "MedLDA: Maximum Margin Supervised Topic Models for Regression and Classification." International Conference on Machine Learning, 2009. doi:10.1145/1553374.1553535Markdown
[Zhu et al. "MedLDA: Maximum Margin Supervised Topic Models for Regression and Classification." International Conference on Machine Learning, 2009.](https://mlanthology.org/icml/2009/zhu2009icml-medlda/) doi:10.1145/1553374.1553535BibTeX
@inproceedings{zhu2009icml-medlda,
title = {{MedLDA: Maximum Margin Supervised Topic Models for Regression and Classification}},
author = {Zhu, Jun and Ahmed, Amr and Xing, Eric P.},
booktitle = {International Conference on Machine Learning},
year = {2009},
pages = {1257-1264},
doi = {10.1145/1553374.1553535},
url = {https://mlanthology.org/icml/2009/zhu2009icml-medlda/}
}