本文研究一类低秩矩阵优化问题,其中惩罚项为目标矩阵奇异值的l_(p)(0<p<1)正则函数.基于半阈值函数在稀疏/低秩恢复问题中的良好性能,本文提出奇异值半阈值(singular value half thresholding,SVHT)算法来求解l_(p)正则矩阵优化...本文研究一类低秩矩阵优化问题,其中惩罚项为目标矩阵奇异值的l_(p)(0<p<1)正则函数.基于半阈值函数在稀疏/低秩恢复问题中的良好性能,本文提出奇异值半阈值(singular value half thresholding,SVHT)算法来求解l_(p)正则矩阵优化问题.SVHT算法的主要迭代利用了子问题的闭式解,但与现有算法不同,其本质上是对目标函数在当前点进行局部1/2近似,而不是局部线性或局部二次近似.通过构造目标函数的Lipschitz和非Lipschitz近似函数,本文证明了SVHT算法生成序列的任意聚点都是问题的一阶稳定点.在数值实验中,利用模拟数据和实际图像数据的低秩矩阵补全问题对SVHT算法进行测试.大量的数值结果表明,SVHT算法对低秩矩阵优化问题在速度、精度和鲁棒性等方面都表现优异.展开更多
A new limited memory symmetric rank one algorithm is proposed. It combines a modified self-scaled symmetric rank one （SSR1） update with the limited memory and nonmonotone line search technique. In this algorithm, th...A new limited memory symmetric rank one algorithm is proposed. It combines a modified self-scaled symmetric rank one （SSR1） update with the limited memory and nonmonotone line search technique. In this algorithm, the descent search direction is generated by inverse limited memory SSR1 update, thus simplifying the computation. Numerical comparison of the algorithm and the famous limited memory BFGS algorithm is given. Comparison results indicate that the new algorithm can process a kind of large-scale unconstrained optimization problems.展开更多
The matrix rank minimization problem arises in many engineering applications. As this problem is NP-hard, a nonconvex relaxation of matrix rank minimization, called the Schatten-p quasi-norm minimization(0 < p <...The matrix rank minimization problem arises in many engineering applications. As this problem is NP-hard, a nonconvex relaxation of matrix rank minimization, called the Schatten-p quasi-norm minimization(0 < p < 1), has been developed to approximate the rank function closely. We study the performance of projected gradient descent algorithm for solving the Schatten-p quasi-norm minimization(0 < p < 1) problem.Based on the matrix restricted isometry property(M-RIP), we give the convergence guarantee and error bound for this algorithm and show that the algorithm is robust to noise with an exponential convergence rate.展开更多
基金Supported by the National Natural Science Foundation of China(10971102)the Natural Science Foundation of Jiangsu Province of China(BK2009398)+1 种基金the Foundation for the Authors of the National Excellent Doctoral Thesis Award of China(200720)Jiangsu Innovation Fund for Doctor of Science(CX07B-027z)
文摘本文研究一类低秩矩阵优化问题,其中惩罚项为目标矩阵奇异值的l_(p)(0<p<1)正则函数.基于半阈值函数在稀疏/低秩恢复问题中的良好性能,本文提出奇异值半阈值(singular value half thresholding,SVHT)算法来求解l_(p)正则矩阵优化问题.SVHT算法的主要迭代利用了子问题的闭式解,但与现有算法不同,其本质上是对目标函数在当前点进行局部1/2近似,而不是局部线性或局部二次近似.通过构造目标函数的Lipschitz和非Lipschitz近似函数,本文证明了SVHT算法生成序列的任意聚点都是问题的一阶稳定点.在数值实验中,利用模拟数据和实际图像数据的低秩矩阵补全问题对SVHT算法进行测试.大量的数值结果表明,SVHT算法对低秩矩阵优化问题在速度、精度和鲁棒性等方面都表现优异.
基金the National Natural Science Foundation of China(10471062)the Natural Science Foundation of Jiangsu Province(BK2006184)~~
文摘A new limited memory symmetric rank one algorithm is proposed. It combines a modified self-scaled symmetric rank one （SSR1） update with the limited memory and nonmonotone line search technique. In this algorithm, the descent search direction is generated by inverse limited memory SSR1 update, thus simplifying the computation. Numerical comparison of the algorithm and the famous limited memory BFGS algorithm is given. Comparison results indicate that the new algorithm can process a kind of large-scale unconstrained optimization problems.
基金supported by National Natural Science Foundation of China(Grant No.11171299)
文摘The matrix rank minimization problem arises in many engineering applications. As this problem is NP-hard, a nonconvex relaxation of matrix rank minimization, called the Schatten-p quasi-norm minimization(0 < p < 1), has been developed to approximate the rank function closely. We study the performance of projected gradient descent algorithm for solving the Schatten-p quasi-norm minimization(0 < p < 1) problem.Based on the matrix restricted isometry property(M-RIP), we give the convergence guarantee and error bound for this algorithm and show that the algorithm is robust to noise with an exponential convergence rate.
