%0 Journal Article
%T 正则化交替方向乘子算法求解非凸不可分离问题<br>Regualized Alternating Direction Method of Multiplies for Nonseparable Nonconvex Problem
%A 胡惠晴
%A 党亚峥
%A 张翼鹏
%J Pure Mathematics
%P 655-665
%@ 2160-7605
%D 2020
%I Hans Publishing
%R 10.12677/PM.2020.107079
%X <p align=\"justify\">
	本文作者提出了一种新的正则化交替方向乘子法来解决非凸优化问题，在不要求正则项严格凸的情况下证明了算法的全局收敛性，在增广拉格朗日函数满足KL性质的条件下，证明了算法的强收敛性，并且通过应用于Lasso模型求解，证明了算法的有效性。<br />
In this paper, we propose a new regularized alternating direction method of multipliers to solve the non-convex optimization problem. We prove the global convergence of the algorithm without requiring the strict convexity of the regular term. Under the condition that the augmented Lagrangian function satisfies KL function, the strong convergence of the algorithm is proved, and the effectiveness of the algorithm is proved by applying it to the Lasso model.
</p>
%K 非凸最优化，正则化，交替方向乘子法，收敛性，Lasso<br>Nonconvex Optimization Problem
%K Regularization
%K Alternating Direction Method of Multiplies
%K Convergence
%K Lasso
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=36719