{"id":"arxiv/2608.00952","url":"https://arxiv.org/abs/2608.00952","tags":["math.NA"],"score":0,"title":"Randomized iterative methods with Polyak step-size for solving generalized absolute value equations","domain":"math.NA","source":"arxiv","authors":["Jiayun Chen","Qiye Zhang","Deren Han","Jiaxin Xie"],"reviews":0,"subject":"math.NA","summary":"In this paper, we systematically incorporate the Polyak step-size into the randomized iterative method to improve its efficiency for solving generalized absolute value equations. In particular, we adopt the Polyak step-size within a stochastic iterative setting where the objective function updates dynamically at every step, unlike the classical Polyak step-size designed for deterministic optimization with fixed objective functions. Consequently, this novel implementation differs from the conventional Polyak scheme and demands a dedicated convergence analysis. We rigorously analyze the…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-08-02","reactions":{}}