A Symmetric Regularized ADMM for Nonconvex Nonseparable Optimization
摘要
In this paper, for the nonconvex nonseparable optimization with linear constraints, a symmetric regularized alternating direction method of multipliers (SRADMM) is proposed. Under this condition that the corresponding augmented Lagrangian function satisfies the Kurdyka–Łojasiewicz (KL) property, we show that the sequence generated by the proposed method converges to the stable point. Some numerical experiments are reported to show that the proposed method is feasible and effective.