https://hackerpaper.kaipfister.de/p/1701.07336/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMThttps://news.ycombinator.com/item?id=22179496urn:hackerpapers:mention:1701.07336:22179496Wed, 29 Jan 2020 11:43:19 GMT<p>If you&#x27;re interested, in the scientific literature this problem is known as &quot;network backboning&quot;. Basically you have all nodes connected to practically all other nodes with weighted edges, and you want to know which are the edges with statistically significant weights. I wrote on this topic [1]. My method [2] basically uses simple counts on edge weights, and then estimates the expected edge weight and its variance using Bayesian priors. It then attaches a t-score or p-value to each edge, and then you can filter out edges with too low t-score. The idea is that weak edges can still be…</p><p><a href="https://hackerpaper.kaipfister.de/p/1701.07336/">Paper and discussion history</a>. A mention is not an endorsement.</p>physics.soc-ph