{"id":"arxiv/2605.11627","url":"https://arxiv.org/abs/2605.11627","tags":["math.OC"],"score":0,"title":"Proximal Limited-Memory Quasi-Newton Methods for Nonsmooth Nonconvex Optimization","domain":"math.OC","source":"arxiv","authors":["Simeon vom Dahl","Alberto De Marchi","Christian Kanzow"],"reviews":0,"subject":"math.OC","summary":"We introduce a proximal limited--memory quasi--Newton scheme for minimizing the sum of a continuously differentiable function and a proper, lower semicontinuous and prox-bounded, possibly nonsmooth, function. Both functions might be nonconvex. The method builds upon the computation of scaled proximal operators and is globalized by adaptively updating a regularization parameter based on a criterion of sufficient decrease. We prove global convergence under mild assumptions and then establish convergence of the entire sequence (with rates) under the Kurdyka--Lojasiewicz property. To efficiently…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-05-12","reactions":{}}