不用特殊转轴的既约梯度方法

无论是Wolfe既约梯度法,还是Zangwill凸单纯形法,在不使用越-韩转轴或类似的转轴运算时,都没得到过令人满意的收敛定理.事实上,那样的收敛定理总是在此非退化假设强很多的不太现实的条件下证得的.本文提出一个新的方法,它介于既约梯度法与凸单纯形法之间.有趣的是,无需特殊的转轴运算,在非退化假设下,我们就     

非线性不等式约束最优化问题一个新的广义既约梯度法

本文借助一种新的求基转轴运算建立了带非线性不等式约束最优化问题的一个新的广义既约梯度法．算法不引入任何松驰变量，以致扩大问题的规模，也不需对约束函数和变量的界预先估计．另一重要特点是方法不再使用隐函数理论确定搜索方向，而是由简单的显式给出．因此方法计算量小，结构简单，便于应用．对于非Ｋ—Ｔ点ｘ，我们构造的方向为可行下降的．本文证明了算法具有全局收敛性．     

一类超线性收敛的既约变尺度法

本文将既约梯度法与Ｈｕａｎｇ族变尺度法相结合，给出标准型线性约束规划问题的一类既约变尺度法．在较温和的假设下，算法具有全局收敛性和超线性收敛速度，最后指出本文算法包含和改进几个己有的有效算法．     

约束优化问题的广义既约型梯度法

本文提出了二类解约束优化问题的广义既约型梯度法,从统一角度研究了投影梯度法和既约梯度法的结构及其全局收敛性.本文结果统一、推广了常见的可行方向法.     

线性等式约束优化的既约预条件共轭梯度路径法

采用既约预条件共轭梯度路径结合非单调技术解线性等式约束的非线性优化问题.基于广义消去法将原问题转化为等式约束矩阵的零空间中的一个无约束优化问题,通过一个增广系统获得既约预条件方程,并构造共轭梯度路径解二次模型,从而获得搜索方向和迭代步长.基于共轭梯度路径的良好性质,在合理的假设条件下,证明了算法不仅具有整体收敛性,而且保持快速的超线性收敛速率.进一步,数值计算表明了算法的可行性和有效性.     

具有线性等式和不等式约束的非线性规划的一个算法

§1.引言既约梯度法是求解非线性规划的一类方法.我们目前只看到约束为线性等式或非线性等式的既约梯度法,对于线性不等式或非线性不等式约束的情形还没有相应的既约梯度法.如果通过松驰变量把线性不等式约束化成线性等式的情形处理,则要增加变量的维数,而这是与既约梯度法的思想背道而驰的.在本文中,我们结合既约梯度法与 Ritter在文献[3]中的思想,对具有线性等式和不等式约束的非线性规划问题给出了一种算法,它保留了既约梯度法降低维数的优点,又简化了 Ritter 在[3]中给出的算法.另外,我们还证明了算法的收敛性.     

可适用于退化情况的既约梯度算法

本文讨论了带有线性约束条件的非线性规划问题,提出了一种可以处理退化情况的既约梯度算法。并在目标函数一阶连续可微的较弱条件下证明了算法的全局收敛性。即证明了算法或在有限步内终止于问题的一个Kuhn—Tucker点,或得到一个点列{x~k},其任一聚点均为问题的Kuhn—Tucker点。     

对既约梯度法的改进

既约梯度法是非线性规划的一类方法,国外多年的实用表明,它是卓有成效的.尤其对于带线性约束的非线性规划问题,既约梯度法的效果被认为较其他方法明显的好.早期的既约梯度法要求较强的假设条件,而且不具有收敛性.为了建立具有收敛性的既约梯度法,已有多篇有关的工作.文献[3]提出了一种特别的转轴算法,在此基础上建立了一种在很弱的条件下具有多敛性的既约梯度法.但[3]中方法仍有可改进之处,     

退化约束的既约变尺度法

既约梯度法是求解线性等式与变量非负约束的非线性规划问题的有效方法,它的优点是降低问题的维数.变尺度方法是求解无约束优化问题的快速方法.文[1]将上述两种方法结合起来,给出了约束非退化并采用精确一维搜索的既约变尺度法,并证明了算法的收敛性与超线性收敛速度.但从计算的实现上来说,必须考虑使用非精确搜索的算法.为了使算法的适应范围更加广泛,也需要放弃约束非退化的假设.本文在满足上述两个要求下给出了退化约束条件下并采用非精确一维搜索的既约变尺度法,证明了算法的全局收敛性与超线性的收敛速度.     

约束极小极大问题的一种既约梯度近似法

利用极大熵方法及有关逼近结果，使之与既约梯度法结合，提出了一种求解极小极大非线性规划问题的近似法，并证明了算法的有关收敛性结果。     

k-均值问题的差分隐私算法综述

$k$-均值问题是机器学习和组合优化领域十分重要的问题。它是经典的NP-难问题, 被广泛的应用于数据挖掘、企业生产决策、图像处理、生物医疗科技等领域。随着时代的发展, 人们越来越注重于个人的隐私保护:在决策通常由人工智能算法做出的情况下, 如何保证尽可能多地从数据中挖掘更多信息,同时不泄露个人隐私。近十年来不断有专家学者研究探索带隐私保护的$k$-均值问题, 得到了许多具有理论指导意义和实际应用价值的结果, 本文主要介绍关于$k$-均值问题的差分隐私算法供读者参考。     

State constrained reachability for stochastic hybrid systems

Many control problems can be formulated as driving a system to reach some target states while avoiding some unwanted states. We study this problem for systems with regime change operating in uncertain environments. Nowadays, it is a common practice to model such systems in the framework of stochastic hybrid system models. In this casting, the problem is formalized as a mathematical problem named state constrained stochastic reachability analysis. In the state constrained stochastic reachability analysis, this probability is computed by imposing a constraint on the system to avoid the unwanted states. The scope of this paper is twofold. First we define and investigate the state constrained reachability analysis in an abstract mathematical setting. We define the problem for a general model of stochastic hybrid systems, and we show that the reach probabilities can be computed as solutions of an elliptic integro-differential equation. Moreover, we extend the problem by considering randomized targets. We approach this extension using stochastic dynamic programming. The second scope is to define a developmental setting in which the state constrained reachability analysis becomes more tractable. This framework is based on multilayer modelling of a stochastic system using hierarchical viewpoints. Viewpoints represent a method originated from software engineering, where a system is described by multiple models created from different perspectives. Using viewpoints, the reach probabilities can be easily computed, or even symbolically calculated. The reach probabilities computed in one viewpoint can be used in another viewpoint for improving the system control. We illustrate this technique for trajectory design.     

并行技术在约束凸规划化问题的对偶算法中的应用

用 Rosen(196 1)的投影梯度的方法求解约束凸规划化问题的对偶问题 ,在计算投影梯度方向时 ,涉及求关于原始变量的最小化问题的最优解 .我们用并行梯度分布算法 (PGD)计算出这一极小化问题的近似解 ,证明近似解可以达到任何给定的精度 ,并说明当精度选取合适时 ,Rosen方法仍然是收敛的     

Three-term conjugate gradient method for the convex optimization problem over the fixed point set of a nonexpansive mapping

Many constrained sets in problems such as signal processing and optimal control can be represented as a fixed point set of a certain nonexpansive mapping, and a number of iterative algorithms have been presented for solving a convex optimization problem over a fixed point set. This paper presents a novel gradient method with a three-term conjugate gradient direction that is used to accelerate conjugate gradient methods for solving unconstrained optimization problems. It is guaranteed that the algorithm strongly converges to the solution to the problem under the standard assumptions. Numerical comparisons with the existing gradient methods demonstrate the effectiveness and fast convergence of this algorithm.     

解等式约束优化的既约Hessian依赖域方法

本文提出了解等式约束优化的一个信赖域方法,该方法以既约Hessian逐步二次规划为基础,它享有信赖域方法与既约Hessian方法的优点,在通常条件下,证明了算法的全局收敛性。     

An affine scaling algorithm for linear programming problems with inequality constraints

A primal, interior point method is developed for linear programming problems for which the linear objective function is to be maximised over polyhedra that are not necessarily in standard form. This algorithm concurs with the affine scaling method of Dikin when the polyhedron is in standard form, and satisfies the usual conditions imposed for using that method. If the search direction is regarded as a function of the current iterate, then it is shown that this function has a unique, continuous extension to the boundary. In fact, on any given face, this extension is just the value the search direction would have for the problem of maximising the objective function over that face. This extension is exploited to prove convergence. The algorithm presented here can be used to exploit such special constraint structure as bounds, ranges, and free variables without increasing the size of the linear programming problem.This paper is in final form and no version of it will be submitted for publication elsewhere.     

A method based on Rayleigh quotient gradient flow for extreme and interior eigenvalue problems

Recently, a continuous method has been proposed by Golub and Liao as an alternative way to solve the minimum and interior eigenvalue problems. According to their numerical results, their method seems promising. This article is an extension along this line. In this article, firstly, we convert an eigenvalue problem to an equivalent constrained optimization problem. Secondly, using the Karush-Kuhn-Tucker conditions of this equivalent optimization problem, we obtain a variant of the Rayleigh quotient gradient flow, which is formulated by a system of differential-algebraic equations. Thirdly, based on the Rayleigh quotient gradient flow, we give a practical numerical method for the minimum and interior eigenvalue problems. Finally, we also give some numerical experiments of our method, the Golub and Liao method, and EIGS (a Matlab implementation for computing eigenvalues using restarted Arnoldi’s method) for some typical eigenvalue problems. Our numerical experiments indicate that our method seems promising for most test problems.     

关于线性规划问题熵障碍对偶法的注记

线性规划是目标优化问题中最常用的模型。关于大规模线性规划问题的有效求解问题一直受到人们的关注。熵障碍对偶法是继内点法之后,又一解线性规划问题的新的算法。本文讨论了熵障碍对偶法的推广形式及其梯度类算法的收敛性。     

Generalized Lagrange multiplier technique for nonlinear programming

Our aim here is to present numerical methods for solving a general nonlinear programming problem. These methods are based on transformation of a given constrained minimization problem into an unconstrained maximin problem. This transformation is done by using a generalized Lagrange multiplier technique. Such an approach permits us to use Newton's and gradient methods for nonlinear programming. Convergence proofs are provided, and some numerical results are given.     

An exact penalty function method for nonlinear mixed discrete programming problems

In this paper, we consider a general class of nonlinear mixed discrete programming problems. By introducing continuous variables to replace the discrete variables, the problem is first transformed into an equivalent nonlinear continuous optimization problem subject to original constraints and additional linear and quadratic constraints. Then, an exact penalty function is employed to construct a sequence of unconstrained optimization problems, each of which can be solved effectively by unconstrained optimization techniques, such as conjugate gradient or quasi-Newton methods. It is shown that any local optimal solution of the unconstrained optimization problem is a local optimal solution of the transformed nonlinear constrained continuous optimization problem when the penalty parameter is sufficiently large. Numerical experiments are carried out to test the efficiency of the proposed method.