Composite optimization for robust rank one bilinear sensing

Composite optimization for robust rank one bilinear sensing
复制标题

稳健的一级双线性传感的复合优化

DOI:
10.1093/imaiai/iaaa027
复制
发表时间:
2020
期刊:
Information and Inference: A Journal of the IMA
影响因子:
--
通讯作者:
Drusvyatskiy, Dmitriy
Drusvyatskiy, Dmitriy
中科院分区:
--
文献类型:
--
作者:
Charisopoulos, Vasileios;Davis, Damek;Díaz, Mateo;Drusvyatskiy, Dmitriy

文献摘要

参考文献

相似文献

我们考虑了从一组可能被噪声破坏的一阶双线性测量中恢复一对向量的任务。最值得注意的是,稳健盲解卷积问题可以用这种方式建模。我们考虑了一阶双线性传感问题的自然非光滑形式,证明了在有利的统计假设下,它的弱凸性、锐性和Lipschitz连续性的模都是维度无关的。即使高达一半的测量被噪声破坏,这种现象也会持续存在。因此,标准算法,如次梯度方法和近似线性方法,当在解的恒定相对误差范围内初始化时,以与维度无关的速度快速收敛。本文提出了一种新的初始化策略,对局部搜索算法进行了补充。初始化过程既被证明是有效的,又对异地测量是稳健的。通过模拟数据和实际数据的数值实验,验证了所提出的理论和方法。
We consider the task of recovering a pair of vectors from a set of rank one bilinear measurements, possibly corrupted by noise. Most notably, the problem of robust blind deconvolution can be modeled in this way. We consider a natural nonsmooth formulation of the rank one bilinear sensing problem and show that its moduli of weak convexity, sharpness and Lipschitz continuity are all dimension independent, under favorable statistical assumptions. This phenomenon persists even when up to half of the measurements are corrupted by noise. Consequently, standard algorithms, such as the subgradient and prox-linear methods, converge at a rapid dimension-independent rate when initialized within a constant relative error of the solution. We complete the paper with a new initialization strategy, complementing the local search algorithms. The initialization procedure is both provably efficient and robust to outlying measurements. Numerical experiments, on both simulated and real data, illustrate the developed theory and methods.
DOI: 10.1109/tsp.2020.3017929
发表时间: 2019-06
影响因子: 5.4
作者:
A. Aghasi;Ali Ahmed;Paul Hand;Babhru Joshi
通讯作者: A. Aghasi;Ali Ahmed;Paul Hand;Babhru Joshi
DOI: 10.1007/s10208-019-09429-9
发表时间: 2017-11
影响因子: 3
作者:
Cong Ma;Kaizheng Wang;Yuejie Chi;Yuxin Chen
通讯作者: Cong Ma;Kaizheng Wang;Yuejie Chi;Yuxin Chen
Banach空间中半凹函数的奇异性
DOI: 10.1007/978-1-4612-1784-8_10
发表时间: 1999
期刊: Renewable Energy
影响因子: 8.7
作者:
P. Albano;P. Cannarsa
通讯作者: P. Cannarsa
DOI: 10.1287/moor.2017.0889
发表时间: 2016-02
期刊: Math. Oper. Res.
影响因子: --
作者:
D. Drusvyatskiy;A. Lewis
通讯作者: D. Drusvyatskiy;A. Lewis
DOI: 10.1109/tpami.2019.2939237
发表时间: 2019-01
影响因子: 23.6
作者:
Yuqian Zhang;Yenson Lau;Han-Wen Kuo;S. Cheung;A. Pasupathy;John Wright
通讯作者: Yuqian Zhang;Yenson Lau;Han-Wen Kuo;S. Cheung;A. Pasupathy;John Wright