ACCAMS: Additive Co-Clustering to Approximate Matrices Succinctly

PDF

Matrix completion and approximation are popular tools to capture a user's preferences for recommendation and to approximate missing data. Instead of using low-rank factorization we take a drastically different approach, based on the simple insight that an additive model of co-clusterings allows one to approximate matrices efficiently. This allows us to build a concise model that, per bit of model learned, significantly beats all factorization approaches to matrix approximation. Even more surprisingly, we find that summing over small co-clusterings is more effective in modeling matrices than classic co-clustering, which uses just one large partitioning of the matrix. Following Occam's razor principle suggests that the simple structure induced by our model better captures the latent preferences and decision making processes present in the real world than classic co-clustering or matrix factorization. We provide an iterative minimization algorithm, a collapsed Gibbs sampler, theoretical guarantees for matrix approximation, and excellent empirical evidence for the efficacy of our approach. We achieve state-of-the-art results on the Netflix problem with a fraction of the model complexity.
Submitted 31 Dec 2014 to Learning [cs.LG]
Published 5 Jan 2015
Subjects: cs.LG stat.ML
Author comments: 22 pages, under review for conference publication
http://arxiv.org/abs/1501.00199
http://arxiv.org/pdf/1501.00199.pdf

0 comments