{"id":"arxiv/2608.21359","url":"https://arxiv.org/abs/2608.21359","tags":["math.OC","cs.AI","cs.LG"],"score":0,"title":"Primal Acceleration of Newton's Method","domain":"math.OC","source":"arxiv","authors":["Nikita Doikov"],"reviews":0,"subject":"math.OC","summary":"We develop a new direct accelerated Newton method for minimizing convex functions with Lipschitz continuous Hessian. The algorithm uses only primal variables and performs just one linear solve per iteration. With a simple predetermined choice of parameters, it achieves the global convergence rate of $O(1/k^3)$ in terms of the functional residual. To the best of our knowledge, this is the first second-order method for this problem class attaining this rate while relying solely on one linear system solve per iteration (without solving auxiliary nonlinear regularized subproblems, such as cubic…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-08-21","reactions":{}}