Efficient MCMC sampling for Bayesian matrix factorization by breaking posterior symmetries

Saibal De; Hadi Salehi; Alex Gorodetsky

Efficient MCMC sampling for Bayesian matrix factorization by breaking posterior symmetries

Saibal De
, Hadi Salehi
, Alex Gorodetsky

University of Michigan

Research output: Contribution to journal › Article › peer-review

Original language	American English
Journal	arXiv preprint arXiv:2006.04295
State	Published - Jun 2020
Externally published	Yes

Disciplines

Engineering
Computational Engineering

Cite this

@article{9b4ba80506d84c1abffe02c9af8f6ba3,

title = "Efficient MCMC sampling for Bayesian matrix factorization by breaking posterior symmetries",

author = "Saibal De and Hadi Salehi and Alex Gorodetsky",

note = "Bayesian low-rank matrix factorization techniques have become an essential tool for relational data analysis and matrix completion. A standard approach is to assign zero-mean Gaussian priors on the columns or rows of factor matrices to create a conjugate system.",

year = "2020",

month = jun,

language = "American English",

journal = "arXiv preprint arXiv:2006.04295",

}

TY - JOUR

T1 - Efficient MCMC sampling for Bayesian matrix factorization by breaking posterior symmetries

AU - De, Saibal

AU - Salehi, Hadi

AU - Gorodetsky, Alex

N1 - Bayesian low-rank matrix factorization techniques have become an essential tool for relational data analysis and matrix completion. A standard approach is to assign zero-mean Gaussian priors on the columns or rows of factor matrices to create a conjugate system.

PY - 2020/6

Y1 - 2020/6

UR - https://arxiv.org/abs/2006.04295

M3 - Article

JO - arXiv preprint arXiv:2006.04295

JF - arXiv preprint arXiv:2006.04295

ER -

Efficient MCMC sampling for Bayesian matrix factorization by breaking posterior symmetries

Disciplines

Other files and links

Cite this