谱HS投影算法求解非线性单调方程组

借助谱梯度法和HS共轭梯度法的结构,建立一种求解非线性单调方程组问题的谱HS投影算法.该算法继承了谱梯度法和共辄梯度法储存量小和计算简单的特征,且不需要任何导数信息,因此它适应于求解大规模非光滑的非线性单调方程组问题.在适当的条件下,证明了该算法的收敛性,并通过数值实验表明了该算法的有效性.     

修正Hestenes-Stiefel共轭梯度算法

本文探讨了Hestenes－Stiefel（HS）共轭梯度算法的收敛性条件．在无充分下降性条件下，证明了一种修正的HS共轭梯度算法的整体收敛性．     

结合Armijo步长搜索的一类新的记忆梯度算法及其收敛特征

对求解无约束规划的超记忆梯度算法中线搜索方向中的参数,给了一个假设条件,从而确定了它的一个新的取值范围,保证了搜索方向是目标函数的充分下降方向,由此提出了一类新的记忆梯度算法.在去掉迭代点列有界和Armijo步长搜索下,讨论了算法的全局收敛性,且给出了结合形如共轭梯度法FR,PR,HS的记忆梯度法的修正形式.数值实验表明,新算法比Armijo线搜索下的FR、PR、HS共轭梯度法和超记忆梯度法更稳定、更有效.     

一个充分下降的改进HS共轭梯度法

共轭梯度法是求解大规模无约束优化问题最有效的方法之一.对HS共轭梯度法参数公式进行改进,得到了一个新公式,并以新公式建立一个算法框架.在不依赖于任何线搜索条件下,证明了由算法框架产生的迭代方向均满足充分下降条件,且在标准Wolfe线搜索条件下证明了算法的全局收敛性.最后,对新算法进行数值测试,结果表明所改进的方法是有效的.     

建立在修正BFGS公式基础上的新的共轭梯度法

共轭梯度法是一类非常重要的用于解决大规模无约束优化问题的方法. 本文通过修正的BFGS公式提出了一个新的共轭梯度方法. 该方法具有不依赖于线搜索的充分下降性. 对于一般的非线性函数, 证明了该方法的全局收敛性. 数值结果表明该方法是有效的.     

修正PRP共轭梯度方法求解无约束最优化问题

基于著名的PRP共轭梯度方法,利用CG_DESCENT共轭梯度方法的结构,本文提出了一种求解大规模无约束最优化问题的修正PRP共轭梯度方法。该方法在每一步迭代中均能够产生一个充分下降的搜索方向,且独立于任何线搜索条件。在标准Wolfe线搜索条件下,证明了修正PRP共轭梯度方法的全局收敛性和线性收敛速度。数值结果展示了修正PRP方法对给定的测试问题是非常有效的。     

修正PRP共轭梯度方法求解无约束最优化问题

基于著名的PRP共轭梯度方法,利用CG_DESCENT共轭梯度方法的结构,本文提出了一种求解大规模无约束最优化问题的修正PRP共轭梯度方法。该方法在每一步迭代中均能够产生一个充分下降的搜索方向,且独立于任何线搜索条件。在标准Wolfe线搜索条件下,证明了修正PRP共轭梯度方法的全局收敛性和线性收敛速度。数值结果展示了修正PRP方法对给定的测试问题是非常有效的。     

一种自适应Dai-Liao共轭梯度法

共轭梯度法是一种解决大规模无约束优化问题的重要方法.本文对Dai-Liao (DL)共轭梯度法的参数进行了研究,提出了一种新的自适应DL共轭梯度法.在适当的条件下,证明了该方法的全局收敛性.数值结果表明,我们的方法对给定的测试问题是有效的.     

凸约束伪单调方程组的无导数投影算法

基于HS共轭梯度法的结构,本文在弱假设条件下建立了一种求解凸约束伪单调方程组问题的迭代投影算法.该算法不需要利用方程组的任何梯度或Jacobian矩阵信息,因此它适合求解大规模问题.算法在每一次迭代中都能产生充分下降方向,且不依赖于任何线搜索条件.特别是,我们在不需要假设方程组满足Lipschitz条件下建立了算法的全局收敛性和R-线收敛速度.数值结果表明,该算法对于给定的大规模方程组问题是稳定和有效的.     

特殊双变量矩阵方程组异类约束解的MCG算法

本文研究了约束矩阵方程问题中异类约束解的迭代算法.利用修正共轭梯度法,求得了特殊双变量线性矩阵方程组的异类约束解,选取特殊的初始矩阵,得到唯一极小范数异类约束解.理论证明和数值算例验证了该方法的有限步收敛性,推广了修正共轭梯度法在求约束矩阵方程问题中的应用范围.     

基于l1-l2范数的块稀疏信号重构

压缩感知(compressed sensing,CS) 是一种全新的信息采集与处理的理论框架,借助信号内在的稀疏性或可压缩性,可以从小规模的线性、非自适应的测量中通过求解非线性优化问题重构原信号.块稀疏信号是一种具有块结构的信号,即信号的非零元是成块出现的.受YIN Peng-hang, LOU Yi-fei, HE Qi等提出的l_1-2范数最小化方法的启发,将基于l₁-l₂范数的稀疏重构算法推广到块稀疏模型,证明了块稀疏模型下l₁-l₂范数的相关性质,建立了基于l₁-l₂范数的块稀疏信号精确重构的充分条件,并通过DCA(difference of convex functions algorithm) 和ADMM(alternating direction method of multipliers)给出了求解块稀疏模型下l₁-l₂范数的迭代方法.数值实验表明,基于l₁-l₂范数的块稀疏重构算法比其他块稀疏重构算法具有更高的重构成功率.     

复合凸优化的快速邻近点算法

在大数据时代,随着数据采集手段的不断提升,大规模复合凸优化问题大量的出现在包括统计数据分析,机器与统计学习以及信号与图像处理等应用中.本文针对大规模复合凸优化问题介绍了一类快速邻近点算法.在易计算的近似准则和较弱的平稳性条件下,本文给出了该算法的全局收敛与局部渐近超线性收敛结果.同时,我们设计了基于对偶原理的半光滑牛顿法来高效稳定求解邻近点算法所涉及的重要子问题.最后,本文还讨论了如何通过深入挖掘并利用复合凸优化问题中由非光滑正则函数所诱导的非光滑二阶信息来极大减少半光滑牛顿算法中求解牛顿线性系统所需的工作量,从而进一步加速邻近点算法.     

Quantitative Uncertainty Principles for the Canonical Fourier—Bessel Transform

Acta Mathematica Sinica, English Series - The aim of this paper is to establish an extension of quantitative uncertainty principles and an algorithm for signal recovery about the essential supports...     

A heuristic approach based on shortest path problems for integrated flight,aircraft, and passenger rescheduling under disruptions

In this paper, we present a heuristic method to solve an airline disruption management problem arising from the ROADEF 2009 challenge. Disruptions perturb an initial flight plan such that some passengers cannot start or conclude their planned trip. The developed algorithm considers passengers and aircraft with the same priority by reassigning passengers and by creating a limited number of flights. The aim is to minimize the cost induced for the airline by the recovery from the disruptions. The algorithm is tested on real-life-based data, as well as on large-scale instances and ranks among the best methods proposed to the challenge in terms of quality, while being efficient in terms of computation time.     

Convex Variational Formulation with Smooth Coupling for Multicomponent Signal Decomposition and Recovery

A convex variational formulation is proposed to solve multicomponent signal processing problems in Hilbert spaces. The cost function consists of a separable term, in which each component is modeled through its own potential, and of a coupling term, in which constraints on linear transformations of the components are penalized with smooth functionals. An algorithm with guaranteed weak convergence to a solution to the problem is provided. Various multicomponent signal decomposition and recovery applications are discussed.     

Dualization of Signal Recovery Problems

In convex optimization, duality theory can sometimes lead to simpler solution methods than those resulting from direct primal analysis. In this paper, this principle is applied to a class of composite variational problems arising in particular in signal recovery. These problems are not easily amenable to solution by current methods but they feature Fenchel–Moreau–Rockafellar dual problems that can be solved by forward-backward splitting. The proposed algorithm produces simultaneously a sequence converging weakly to a dual solution, and a sequence converging strongly to the primal solution. Our framework is shown to capture and extend several existing duality-based signal recovery methods and to be applicable to a variety of new problems beyond their scope.     

Optimum signal and image recovery by the method of alternating projections in fractional Fourier domains

This paper presents a signal and image recovery scheme by the method of alternating projections onto convex sets in optimum fractional Fourier domains. It is shown that the fractional Fourier domain order with minimum bandwidth is the optimum fractional Fourier domain for the method employing alternating projections in signal recovery problems. Following the estimation of optimum fractional Fourier transform orders, incomplete signal is projected onto different convex sets consecutively to restore the missing part. Using a priori information in optimum fractional Fourier domains, superior results are obtained compared to the conventional Fourier domain restoration. The algorithm is tested on 1-D linear frequency modulated signals, real biological data and 2-D signals presenting chirp-type characteristics. Better results are obtained in the matched fractional Fourier domain, compared to not only the conventional Fourier domain restoration, but also other fractional Fourier domains.     

Infeasible Predictor-Corrector Interior-Point Method Applied to Image Restoration in the Presence of Noise

Image recovery problems can be solved using optimization techniques. They lead often to the solution of either a large-scale convex quadratic program or equivalently a nondifferentiable minimization problem. To solve the quadratic program, we use an infeasible predictor-corrector interior-point method, presented in the more general framework of monotone LCP. The algorithm has polynomial complexity and it converges with asymptotic quadratic rate. When implementing the method to recover images, we take advantage of the underlying sparsity of the problem. We obtain good performances, that we assess by comparing the method with a variable-metric proximal bundle algorithm applied to the solution of equivalent nonsmooth problem.     

求解低秩矩阵恢复问题的非单调交替梯度方向法

低秩矩阵恢复问题作为一类在图像处理和信号数据分析等领域中都十分重要的问题已被广泛研究.本文在交替方向算法的框架下,应用非单调技术,提出一种求解低秩矩阵恢复问题的新算法.该算法在每一步迭代过程中,首先利用一步带有变步长梯度算法同时更新低秩部分的两块变量,然后采用非单调技术更新稀疏部分的变量.在一定的假设条件下,本文证明了新算法的全局收敛性.最后通过解决随机低秩矩阵恢复问题和视频前景背景分离的实例验证新算法的有效性,同时也显示非单调技术极大改善了算法的效率.     

A GREEDY ALGORITHM FOR SPARSE PRECISION MATRIX APPROXIMATION

Precision matrix estimation is an important problem in statistical data analysis.This paper proposes a sparse precision matrix estimation approach,based on CLIME estimator and an efficient algorithm GISSρ that was originally proposed for l1 sparse signal recov-ery in compressed sensing.The asymptotic convergence rate for sparse precision matrix estimation is analyzed with respect to the new stopping criteria of the proposed GISSρ algorithm.Finally,numerical comparison of GISSρ with other sparse recovery algorithms,such as ADMM and HTP in three settings of precision matrix estimation is provided and the numerical results show the advantages of the proposed algorithm.