dedupe icon indicating copy to clipboard operation
dedupe copied to clipboard

Published 20 hours ago • verified publisher icon dedupeio

partially supervised classification

Open fgregg opened this issue 3 years ago • 1 comments

PU Learning looks like it might be a great fit for record-linkage problems? https://www.cs.uic.edu/~liub/NSF/PSC-IIS-0307239.html

fgregg avatar Jan 24 '22 15:01 fgregg

splink seems to somehow guess which records are matches using some sort of unsupervised Expectation Maximisation algorithm (that it does not explain very well and I couldn't find a good explanation for anywhere...)

NickCrews avatar Sep 24 '22 20:09 NickCrews