Sinkhorn Permutation Variational Marginal Inference

Gonzalo Mena; Erdem Varol; Amin Nejatbakhsh; Eviatar Yemini; Liam Paninski

Sinkhorn Permutation Variational Marginal Inference

Gonzalo Mena, Erdem Varol, Amin Nejatbakhsh, Eviatar Yemini, Liam Paninski

Research output: Contribution to journal › Conference article › peer-review

Abstract

We address the problem of marginal inference for an exponential family defined over the set of permutation matrices. This problem is known to quickly become intractable as the size of the permutation increases, since its involves the computation of the permanent of a matrix, a #P-hard problem. We introduce Sinkhorn variational marginal inference as a scalable alternative, a method whose validity is ultimately justified by the so-called Sinkhorn approximation of the permanent. We demonstrate the effectiveness of our method in the problem of probabilistic identification of neurons in the worm C.elegans.

Original language	English (US)
Journal	Proceedings of Machine Learning Research
Volume	118
State	Published - 2019
Event	2nd Symposium on Advances in Approximate Bayesian Inference, AABI 2019 - Vancouver, Canada Duration: Dec 8 2019 → …

ASJC Scopus subject areas

Artificial Intelligence
Software
Control and Systems Engineering
Statistics and Probability

Cite this

@article{cf8bcbecb35e4bc4a813914d3ee5c0cf,

title = "Sinkhorn Permutation Variational Marginal Inference",

abstract = "We address the problem of marginal inference for an exponential family defined over the set of permutation matrices. This problem is known to quickly become intractable as the size of the permutation increases, since its involves the computation of the permanent of a matrix, a #P-hard problem. We introduce Sinkhorn variational marginal inference as a scalable alternative, a method whose validity is ultimately justified by the so-called Sinkhorn approximation of the permanent. We demonstrate the effectiveness of our method in the problem of probabilistic identification of neurons in the worm C.elegans.",

author = "Gonzalo Mena and Erdem Varol and Amin Nejatbakhsh and Eviatar Yemini and Liam Paninski",

note = "Publisher Copyright: {\textcopyright} G. Mena & E.V.A.N.E.Y.L. Paninski.; 2nd Symposium on Advances in Approximate Bayesian Inference, AABI 2019 ; Conference date: 08-12-2019",

year = "2019",

language = "English (US)",

volume = "118",

journal = "Proceedings of Machine Learning Research",

issn = "2640-3498",

}

TY - JOUR

T1 - Sinkhorn Permutation Variational Marginal Inference

AU - Mena, Gonzalo

AU - Varol, Erdem

AU - Nejatbakhsh, Amin

AU - Yemini, Eviatar

AU - Paninski, Liam

N1 - Publisher Copyright: © G. Mena & E.V.A.N.E.Y.L. Paninski.

PY - 2019

Y1 - 2019

N2 - We address the problem of marginal inference for an exponential family defined over the set of permutation matrices. This problem is known to quickly become intractable as the size of the permutation increases, since its involves the computation of the permanent of a matrix, a #P-hard problem. We introduce Sinkhorn variational marginal inference as a scalable alternative, a method whose validity is ultimately justified by the so-called Sinkhorn approximation of the permanent. We demonstrate the effectiveness of our method in the problem of probabilistic identification of neurons in the worm C.elegans.

AB - We address the problem of marginal inference for an exponential family defined over the set of permutation matrices. This problem is known to quickly become intractable as the size of the permutation increases, since its involves the computation of the permanent of a matrix, a #P-hard problem. We introduce Sinkhorn variational marginal inference as a scalable alternative, a method whose validity is ultimately justified by the so-called Sinkhorn approximation of the permanent. We demonstrate the effectiveness of our method in the problem of probabilistic identification of neurons in the worm C.elegans.

UR - http://www.scopus.com/inward/record.url?scp=85161267405&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85161267405&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:85161267405

SN - 2640-3498

VL - 118

JO - Proceedings of Machine Learning Research

JF - Proceedings of Machine Learning Research

T2 - 2nd Symposium on Advances in Approximate Bayesian Inference, AABI 2019

Y2 - 8 December 2019

ER -

Sinkhorn Permutation Variational Marginal Inference

Abstract

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this