Computing expectations and marginal likelihoods for permutations
Computing expectations and marginal likelihoods for permutations
This paper demonstrates how we can re-purpose sophisticated algorithms from a range of fields to help us compute expected permutations and marginal likelihoods. The results are of particular use in the fields of record linkage or identity resolution, where we are interested in finding pairs of records across data sets that refer to the same individual. All calculations discussed can be reproduced with the accompanying R package expperm.
identity resolution, linkage error, matrix permanent, object tracking, random permutation
871-891
Powell, Ben
866da689-847f-4b75-9150-cd8149684368
Smith, Paul A.
a2548525-4f99-4baf-a4d0-2b216cce059c
1 June 2020
Powell, Ben
866da689-847f-4b75-9150-cd8149684368
Smith, Paul A.
a2548525-4f99-4baf-a4d0-2b216cce059c
Powell, Ben and Smith, Paul A.
(2020)
Computing expectations and marginal likelihoods for permutations.
Computational Statistics, 35 (2), .
(doi:10.1007/s00180-019-00901-2).
Abstract
This paper demonstrates how we can re-purpose sophisticated algorithms from a range of fields to help us compute expected permutations and marginal likelihoods. The results are of particular use in the fields of record linkage or identity resolution, where we are interested in finding pairs of records across data sets that refer to the same individual. All calculations discussed can be reproduced with the accompanying R package expperm.
Text
Powell_Smith_compstat_v2 final
- Accepted Manuscript
Text
Powell-Smith 2020 Article Computing Expectations And Margin
- Version of Record
More information
Accepted/In Press date: 5 June 2019
e-pub ahead of print date: 12 June 2019
Published date: 1 June 2020
Additional Information:
Publisher Copyright:
© 2019, The Author(s).
Keywords:
identity resolution, linkage error, matrix permanent, object tracking, random permutation
Identifiers
Local EPrints ID: 430942
URI: http://eprints.soton.ac.uk/id/eprint/430942
ISSN: 0943-4062
PURE UUID: 74940923-fb06-455e-9b33-f65db5bbbb54
Catalogue record
Date deposited: 17 May 2019 16:30
Last modified: 16 Mar 2024 07:50
Export record
Altmetrics
Contributors
Author:
Ben Powell
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics