Journal article icon

Journal article

Gibbs flow for approximate transport with applications to Bayesian computation

Abstract:

Let πœ‹0 and πœ‹1 be two distributions on the Borel space (ℝ𝑑,𝔅(ℝ𝑑)) . Any measurable function 𝑇:ℝ𝑑→ℝ𝑑 such that π‘Œ=𝑇(𝑋)βˆΌπœ‹1 if π‘‹βˆΌπœ‹0 is called a transport map from πœ‹0 to πœ‹1 . For any πœ‹0 and πœ‹1 , if one could obtain an analytical expression for a transport map from πœ‹0 to πœ‹1 , then this could be straightforwardly applied to sample from any distribution. One would map draws from an easy‐to‐sample distribution πœ‹0 to the target distribution πœ‹1 using this transport map. Although i...

Expand abstract
Publication status:
Published
Peer review status:
Peer reviewed

Actions


Access Document


Files:
Publisher copy:
10.1111/rssb.12404

Authors


More by this author
Institution:
University of Oxford
Division:
MPLS
Department:
Statistics
Role:
Author
More from this funder
Name:
Engineering and Physical Sciences Research Council
Grant:
56726
EP/K000276/1
Publisher:
Wiley
Journal:
Journal of the Royal Statistical Society: Series B (Statistical Methodology) More from this journal
Volume:
83
Issue:
1
Pages:
156-187
Publication date:
2021-01-20
Acceptance date:
2020-10-20
DOI:
EISSN:
1467-9868
ISSN:
1369-7412
Language:
English
Keywords:
Pubs id:
1138332
Local pid:
pubs:1138332
Deposit date:
2020-10-20

Terms of use


Views and Downloads






If you are the owner of this record, you can report an update to it here: Report update to this record

TO TOP