{"id":"arxiv/2608.12665","url":"https://arxiv.org/abs/2608.12665","tags":["math.OC","cs.LG","stat.ML"],"score":0,"title":"A Local-Linearly Convergent Algorithm for Nonconvex Equality-Constrained Optimization","domain":"math.OC","source":"arxiv","authors":["Frank E. Curtis","Lingjun Guo","Daniel P. Robinson"],"reviews":0,"subject":"math.OC","summary":"For solving nonconvex equality-constrained optimization problems, a recent Gradient-Eigenstep Algorithm by Goyens et al.~is an iteration-efficient approach, based on minimizing Fletcher's augmented Lagrangian function, for finding an approximate second-order stationary point from an arbitrary starting point. In this paper, the analysis of this algorithm is extended, offering a two-fold contribution. First, it is shown that a local-linear rate of convergence can be obtained by this method if it is initiated sufficiently close to a strong second-order stationary point and employs a…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-08-12","reactions":{}}