Entropic optimal transport is maximum-likelihood deconvolution
Name
1809.05572.pdf
Description
Submitted version
Size
163.25 KB
Format
Adobe PDF
Checksum (MD5)
0c98dfe0d345fca2c41fba17dc1495e6
Author(s)
Rigollet, Philippe
Weed, Jonathan
Date Issued
November 2018
Journal
Comptes Rendus Mathematique
Publisher
Elsevier BV
Citation
Rigollet, Philippe and Jonathan Weed. "Entropic optimal transport is maximum-likelihood deconvolution." Comptes Rendus Mathematique 356, 11-12 (November 2018): 1228-1235 © 2018 Académie des sciences
Version
Original manuscript
Abstract
We give a statistical interpretation of entropic optimal transport by showing that performing maximum-likelihood estimation for Gaussian deconvolution corresponds to calculating a projection with respect to the entropic optimal transport distance. This structural result gives theoretical support for the wide adoption of these tools in the machine learning community.
MIT Department
Massachusetts Institute of Technology. Department of Mathematics
Terms of Use
Creative Commons Attribution-NonCommercial-NoDerivs License
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1016/j.crma.2018.10.010
