Combinatorial Topic Models using Small-Variance Asymptotics
Abstract
Topic models have emerged as fundamental tools in unsupervised machine learning. Most modern topic modeling algorithms take a probabilistic view and derive inference algorithms based on Latent Dirichlet Allocation (LDA) or its variants. In contrast, we study topic modeling as a combinatorial optimization problem, and propose a new objective function derived from LDA by passing to the small-variance limit. We minimize the derived objective by using ideas from combinatorial optimization, which results in a new, fast, and high-quality topic modeling algorithm. In particular, we show that our results are competitive with popular LDA-based topic modeling approaches, and also discuss the (dis)similarities between our approach and its probabilistic counterparts.
Cite
@article{arxiv.1604.02027,
title = {Combinatorial Topic Models using Small-Variance Asymptotics},
author = {Ke Jiang and Suvrit Sra and Brian Kulis},
journal= {arXiv preprint arXiv:1604.02027},
year = {2016}
}
Comments
19 pages