Current browse context:
cs.LG
Change to browse by:
References & Citations
Computer Science > Machine Learning
Title: Spectral Methods for Correlated Topic Models
(Submitted on 30 May 2016 (v1), last revised 13 Nov 2016 (this version, v5))
Abstract: In this paper, we propose guaranteed spectral methods for learning a broad range of topic models, which generalize the popular Latent Dirichlet Allocation (LDA). We overcome the limitation of LDA to incorporate arbitrary topic correlations, by assuming that the hidden topic proportions are drawn from a flexible class of Normalized Infinitely Divisible (NID) distributions. NID distributions are generated through the process of normalizing a family of independent Infinitely Divisible (ID) random variables. The Dirichlet distribution is a special case obtained by normalizing a set of Gamma random variables. We prove that this flexible topic model class can be learned via spectral methods using only moments up to the third order, with (low order) polynomial sample and computational complexity. The proof is based on a key new technique derived here that allows us to diagonalize the moments of the NID distribution through an efficient procedure that requires evaluating only univariate integrals, despite the fact that we are handling high dimensional multivariate moments. In order to assess the performance of our proposed Latent NID topic model, we use two real datasets of articles collected from New York Times and Pubmed. Our experiments yield improved perplexity on both datasets compared with the baseline.
Submission history
From: Forough Arabshahi [view email][v1] Mon, 30 May 2016 00:32:11 GMT (1263kb)
[v2] Tue, 31 May 2016 14:30:11 GMT (1263kb)
[v3] Sun, 5 Jun 2016 08:27:34 GMT (1264kb)
[v4] Sat, 20 Aug 2016 01:44:30 GMT (1264kb)
[v5] Sun, 13 Nov 2016 20:24:02 GMT (1278kb)
Link back to: arXiv, form interface, contact.