Forward-backward truncated Newton methods for convex composite optimization

作者: Alberto Bemporad , Lorenzo Stella , Panagiotis Patrinos

DOI:

关键词:

摘要: This paper proposes two proximal Newton-CG methods for convex nonsmooth optimization problems in composite form. The algorithms are based on aa reformulation of the original nonsmooth problem as the unconstrained minimization of a continuously differentiable function, namely the forward-backward envelope (FBE). The first algorithm is based on a standard line search strategy, whereas the second one combines the global efficiency estimates of the corresponding first-order methods, while achieving fast asymptotic …

参考文章(56)
Robert Mifflin, Liqun Qi, Defeng Sun, Properties of the Moreau-Yosida regularization of a piecewise C2 convex function Mathematical Programming. ,vol. 84, pp. 269- 281 ,(1999) , 10.1007/S10107980029A
Patrick L. Combettes, Heinz H. Bauschke, Convex Analysis and Monotone Operator Theory in Hilbert Spaces ,(2011)
信子 相良, A Trust Region Method for Nonsmooth Convex Optimization 経営総合科学. pp. 113- 133 ,(2004)
Frank H. Clarke, Optimization and nonsmooth analysis ,(1983)
N. Yamashita, K. Taji, M. Fukushima, Unconstrained Optimization Reformulations of Variational Inequality Problems Journal of Optimization Theory and Applications. ,vol. 92, pp. 439- 456 ,(1997) , 10.1023/A:1022660704427
M. Fazel, H. Hindi, S.P. Boyd, A rank minimization heuristic with application to minimum order system approximation american control conference. ,vol. 6, pp. 4734- 4739 ,(2001) , 10.1109/ACC.2001.945730
Fanwen Meng, Defeng Sun, Gongyun Zhao, Semismoothness of solutions to generalized equations and the Moreau-Yosida regularization Mathematical Programming. ,vol. 104, pp. 561- 581 ,(2005) , 10.1007/S10107-005-0629-9
Jasson D. M. Rennie, Nathan Srebro, Fast maximum margin matrix factorization for collaborative prediction Proceedings of the 22nd international conference on Machine learning - ICML '05. pp. 713- 719 ,(2005) , 10.1145/1102351.1102441