Novel reformulations and efficient algorithms for the generalized trust region subproblem · HackerPapers discussionshttps://hackerpaper.kaipfister.de/p/1707.08706/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMTNovel reformulations and efficient algorithms for the generalized trust region subproblem — mentioned by LolWolfhttps://news.ycombinator.com/item?id=18443248urn:hackerpapers:mention:1707.08706:18443248Tue, 13 Nov 2018 18:14:59 GMT<p>&gt; that reference you can&#x27;t find right now seems rather pertinent? Here it is: https://arxiv.org/pdf/1707.08706.pdf (This isn&#x27;t quite the one I was thinking of, so I&#x27;ll dig a little deeper, but it covers the idea). Some slightly more technical conditions have to hold in order to have vanilla GD work (since the function is non-differentiable at points), but a (very!) slightly generalized variant for convex functions—sub-gradient descent—works. &gt; for an arbitrary non-convex function. Sure, but this is also obvious since it is NP-hard to reach global optima in arbitrary non-convex problems.…</p><p><a href="https://hackerpaper.kaipfister.de/p/1707.08706/">Paper and discussion history</a>. A mention is not an endorsement.</p>math.OC