<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Survey of Nearest Neighbor Techniques · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/1007.0085/</link><description>Collected HN comments and submissions linking this paper.</description><language>en</language><lastBuildDate>Mon, 02 Nov 2026 11:16:22 GMT</lastBuildDate><atom:link href="https://hackerpaper.kaipfister.de/p/1007.0085/feed.xml" rel="self" type="application/rss+xml" /><item><title>Survey of Nearest Neighbor Techniques — mentioned by papaf</title><link>https://news.ycombinator.com/item?id=14099372</link><guid isPermaLink="false">urn:hackerpapers:mention:1007.0085:14099372</guid><pubDate>Wed, 12 Apr 2017 16:44:18 GMT</pubDate><description>&lt;p&gt;I have no experience with fuzzy matching but it sounds a lot like finding Nearest Neighbours to me. It could be techniques used with Nearest Neighbours also help with your problem: Multidimensional scaling is a nice way to visualise distances between objects: https://en.wikipedia.org/wiki/Multidimensional_scaling A review of Nearest Neighour techniques might be interesting: https://arxiv.org/abs/1007.0085 Also, if you stick with just a similarity score, something like a ROC curve can help to pick a cutoff: https://en.wikipedia.org/wiki/Receiver_operating_characteris...&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/1007.0085/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.CV</category></item></channel></rss>
