We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.LG

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: Beyond LDA: A Unified Framework for Learning Latent Normalized Infinitely Divisible Topic Models through Spectral Methods

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 learnt 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.
Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML)
Cite as: arXiv:1605.09080 [cs.LG]
  (or arXiv:1605.09080v2 [cs.LG] for this version)

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.