Constrained Minimax Problems Research Articles

A SSLE-Type Algorithm of Quasi-Strongly Sub-Feasible Directions for Inequality Constrained Minimax Problems

A SSLE-Type Algorithm of Quasi-Strongly Sub-Feasible Directions for Inequality Constrained Minimax Problems

Anderson Acceleration for Nonsmooth Fixed Point Problems

We give new convergence results of Anderson acceleration for the composite $\max$ fixed point problem. We prove that Anderson(1) and EDIIS(1) are q-linear convergent with a smaller q-factor than existing q-factors. Moreover, we propose a smoothing approximation of the composite max function in the contractive fixed point problem. We show that the smoothing approximation is a contraction mapping with the same fixed point as the composite $\max$ fixed point problem. Our results rigorously confirm that the nonsmoothness does not affect the convergence rate of the Anderson acceleration method when we use the proposed smoothing approximation for the composite $\max$ fixed point problem. Numerical results for constrained minimax problems, complementarity problems, and nonsmooth differential equations are presented to show the efficiency and good performance of the proposed Anderson acceleration method with smoothing approximation.

A proximal-projection partial bundle method for convex constrained minimax problems

In this paper, we propose a partial bundle method for a convex constrained minimax problem where the objective function is expressed as maximum of finitely many convex (not necessarily differentiable) functions. To avoid complete evaluation of all component functions of the objective, a partial cutting-planes model is adopted instead of the traditional one. Based on the proximal-projection idea, at each iteration, an unconstrained proximal subproblem is solved first to generate an aggregate linear model of the objective function, and then another subproblem based on this model is solved to obtain a trial point. Moreover, a new descent test criterion is proposed, which can not only simplify the presentation of the algorithm, but also improve the numerical performance significantly. An explicit upper bound for the number of bundle resets is also derived. Global convergence of the algorithm is established, and some preliminary numerical results show that our method is very encouraging.

Augmented Lagrangian functions for cone constrained optimization: the existence of global saddle points and exact penalty property

In the article we present a general theory of augmented Lagrangian functions for cone constrained optimization problems that allows one to study almost all known augmented Lagrangians for cone constrained programs within a unified framework. We develop a new general method for proving the existence of global saddle points of augmented Lagrangian functions, called the localization principle. The localization principle unifies, generalizes and sharpens most of the known results on existence of global saddle points, and, in essence, reduces the problem of the existence of saddle points to a local analysis of optimality conditions. With the use of the localization principle we obtain first necessary and sufficient conditions for the existence of a global saddle point of an augmented Lagrangian for cone constrained minimax problems via both second and first order optimality conditions. In the second part of the paper, we present a general approach to the construction of globally exact augmented Lagrangian functions. The general approach developed in this paper allowed us not only to sharpen most of the existing results on globally exact augmented Lagrangians, but also to construct first globally exact augmented Lagrangian functions for equality constrained optimization problems, for nonlinear second order cone programs and for nonlinear semidefinite programs. These globally exact augmented Lagrangians can be utilized in order to design new superlinearly (or even quadratically) convergent optimization methods for cone constrained optimization problems.

An Attack–Defense Model with Inhomogeneous Resources of the Opponents

Germeier’s attack–defense model is generalized by taking into account the inhomogeneity of the defense resources. It is based on Germeier’s generalized equalization principle, which in the general case of inhomogeneous resources leads to convex constrained minimax problems, which can be solved using the subgradient ascent method.

Honey Bee Mating Optimization with Nelder-Mead for Constrained Optimization, Integer Programming and Minimax Problems

Honey Bee Mating Optimization with Nelder-Mead for Constrained Optimization, Integer Programming and Minimax Problems

An improved partial bundle method for linearly constrained minimax problems

In this paper, we propose an improved partial bundle method for solving linearly constrained minimax problems. In order to reduce the number of component function evaluations, we utilize a partial cutting-planes model to substitute for the traditional one. At each iteration, only one quadratic programming subproblem needs to be solved to obtain a new trial point. An improved descent test criterion is introduced to simplify the algorithm. The method produces a sequence of feasible trial points, and ensures that the objective function is monotonically decreasing on the sequence of stability centers. Global convergence of the algorithm is established. Moreover, we utilize the subgradient aggregation strategy to control the size of the bundle and therefore overcome the difficulty of computation and storage. Finally, some preliminary numerical results show that the proposed method is effective.

Second-Order Analysis of Piecewise Linear Functions with Applications to Optimization and Stability

This paper is devoted to second-order variational analysis of a rather broad class of extended-real-valued piecewise liner functions and their applications to various issues of optimization and stability. Based on our recent explicit calculations of the second-order subdifferential for such functions, we establish relationships between nondegeneracy and second-order qualification for fully amenable compositions involving piecewise linear functions. We then provide a second-order characterization of full stable local minimizers in composite optimization and constrained minimax problems.

A feasible descent bundle method for inequality constrained Minimax problems

In this paper, a feasible descent bundle method for solving inequality constrained Minimax problems is proposed. The main features of the method are: (1) By using the subgradients of functions and the idea of bundle method, it does not assume that the component functions of the original problem are smooth; (2) by adopting the technique of partial cutting-planes model, at each null step, only the function value and subgradient of one component function are needed to generate the new cutting plane, and therefore the computational cost is reduced effectively; (3) it can ensure the feasibility of the serious iterates and the decsent of the objective function; (4) by introducing the technique of subgradient aggregation to aggregate the subgradients in the bundle, the difficulty of numerical calculation and storage is overcome; (5) the algorithm is proved to be globally convergent, and the preliminary numerical results show that the proposed algorithm is effective.

Primal-dual algorithms for total variation based image restoration under Poisson noise

We consider the problem of restoring images corrupted by Poisson noise. Under the framework of maximum a posteriori estimator, the problem can be converted into a minimization problem where the objective function is composed of a Kullback-Leibler (KL)-divergence term for the Poisson noise and a total variation (TV) regularization term. Due to the logarithm function in the KL-divergence term, the non-differentiability of TV term and the positivity constraint on the images, it is not easy to design stable and efficiency algorithm for the problem. Recently, many researchers proposed to solve the problem by alternating direction method of multipliers (ADMM). Since the approach introduces some auxiliary variables and requires the solution of some linear systems, the iterative procedure can be complicated. Here we formulate the problem as two new constrained minimax problems and solve them by Chambolle-Pocks first order primal-dual approach. The convergence of our approach is guaranteed by their theory. Comparing with ADMM approaches, our approach requires about half of the auxiliary variables and is matrix-inversion free. Numerical results show that our proposed algorithms are efficient and outperform the ADMM approach.

New active set identification for general constrained optimization and minimax problems

The purpose of this paper is to discuss the problem of identifying the active constraints for general constrained nonlinear programming and constrained minimax problems at an isolated local solution. Facchinei, Fischer and Kanzow [4] proposed an effective technique which can identify the active set in a neighborhood of an isolated local solution for nonlinear programming, and Han, Jian and Li [5] improved and extended this to minimax problems. In this work, a new active constraint identification set is constructed, not only is it tighter than the previous two identification sets, but also it can be used effectively in penalty algorithms. Without strict complementarity and linear independence, it is shown that the new identification technique can accurately identify the active constraints of nonlinear programming and constrained minimax problems. Some numerical examples illustrate the identification technique.

A New Objective Penalty Function Approach for Solving Constrained Minimax Problems

In this paper, a new objective penalty function approach is proposed for solving minimax programming problems with equality and inequality constraints. This new objective penalty function combines the objective penalty and constraint penalty. By the new objective penalty function, a constrained minimax problem is converted to minimizations of a sequence of continuously differentiable functions with a simple box constraint. One can thus apply any efficient gradient minimization methods to solve the minimizations with box constraint at each step of the sequence. Some relationships between the original constrained minimax problem and the corresponding minimization problems with box constraint are established. Based on these results, an algorithm for finding a global solution of the constrained minimax problems is proposed by integrating the particular structure of minimax problems and its global convergence is proved under some conditions. Furthermore, an algorithm is developed for finding a local solution of the constrained minimax problems, with its convergence proved under certain conditions. Preliminary results

Improved Filter-SQP Algorithm with Active Set for Constrained Minimax Problems

An improved filter-SQP algorithm with active set for constrained finite minimax problems is proposed. Firstly, an active constraint subset is obtained by a pivoting operation procedure. Then, a new quadratic programming (QP) subproblem is constructed based on the active constraint subset. The main search directiondkis obtained by solving this (QP) subproblem which is feasible at per iteration point and need not to consider the penalty function by using the filter technique. Under some suitable conditions, the global convergence of our algorithm is established. Finally, some numerical results are reported to show the effectiveness of the proposed algorithm.

Superlinearly Convergent Norm-Relaxed SQP Method Based on Active Set Identification and New Line Search for Constrained Minimax Problems

In this paper, the minimax problems with inequality constraints are discussed, and an alternative fast convergent method for the discussed problems is proposed. Compared with the previous work, the proposed method has the following main characteristics. First, the active set identification which can reduce the scale and the computational cost is adopted to construct the direction finding subproblems. Second, the master direction and high-order correction direction are computed by solving a new type of norm-relaxed quadratic programming subproblem and a system of linear equations, respectively. Third, the step size is yielded by a new line search which combines the method of strongly sub-feasible direction with the penalty method. Fourth, under mild assumptions without any strict complementarity, both the global convergence and rate of superlinear convergence can be obtained. Finally, some numerical results are reported.

A superlinearly convergent norm-relaxed method of quasi-strongly sub-feasible direction for inequality constrained minimax problems

In this paper, nonlinear minimax problems with inequality constraints are discussed. Combined the norm-relaxed SQP method with the idea of strongly sub-feasible directions method, a new method of quasi-strongly sub-feasible directions (MQSSFD) with arbitrary initial point for the discussed problems is presented. At each iteration of the proposed algorithm, an improved search direction is obtained by solving a quadratic program (QP) which always has a solution, and a high-order correction direction is yielded via a system of linear equations (SLE) to avoid the Maratos effect. After finite iterations, the iteration point always get into the feasible set by introducing a new non-monotone curve search. Under some mild conditions including the weak Mangasarian–Fromovitz constraint qualification (MFCQ), the proposed algorithm possesses global convergence, and the superlinear convergence is obtained without the strict complementarity. Finally, some elementary numerical experiments are implemented and reported.

Simple Sequential Quadratically Constrained Quadratic Programming Feasible Algorithm with Active Identification Sets for Constrained Minimax Problems

In this paper, the nonlinear minimax problems with inequality constraints are discussed. Based on the idea of simple sequential quadratically constrained quadratic programming algorithm for smooth constrained optimization, an alternative algorithm for solving the discussed problems is proposed. Unlike the previous work, at each iteration, a feasible direction of descent called main search direction is obtained by solving only one subprogram which is composed of a convex quadratic objective function and simple quadratic inequality constraints without the second derivatives of the constrained functions. Then a high-order correction direction used to avoid the Maratos effect is computed by updating the main search direction with a system of linear equations. The proposed algorithm possesses global convergence under weak Mangasarian–Fromovitz constraint qualification and superlinear convergence under suitable conditions with the upper-level strict complementarity. At last, some preliminary numerical results are reported.

A hybrid algorithm for linearly constrained minimax problems

Many real life problems can be stated as a minimax problem, such as economics, finance, management, engineering and other fields, which demonstrate the importance of having reliable methods to tackle minimax problems. In this paper, an algorithm for linearly constrained minimax problems is presented in which we combine the trust-region methods with the line-search methods and curve-search methods. By means of this hybrid technique, it avoids possibly solving the trust-region subproblems many times, and make better use of the advantages of different methods. Under weaker conditions, the global and superlinear convergence are achieved. Numerical experiments show that the new algorithm is robust and efficient.

A Neural Network Model for Equality and Inequality Constrained Minimax Problems

A Neural Network Model for Equality and Inequality Constrained Minimax Problems

A nonlinear augmented Lagrangian for constrained minimax problems

A novel smooth nonlinear augmented Lagrangian for solving minimax problems with inequality constraints, is proposed in this paper, which has the positive properties that the classical Lagrangian and the penalty function fail to possess. The corresponding algorithm mainly consists of minimizing the nonlinear augmented Lagrangian function and updating the Lagrange multipliers and controlling parameter. It is demonstrated that the algorithm converges Q-superlinearly when the controlling parameter is less than a threshold under the mild conditions. Furthermore, the condition number of the Hessian of the nonlinear augmented Lagrangian function is studied, which is very important for the efficiency of the algorithm. The theoretical results are validated further by the preliminary numerical experiments for several testing problems reported at last, which show that the nonlinear augmented Lagrangian is promising.

On the accurate identification of active set for constrained minimax problems

In this paper, the problem of identifying the active constraints for constrained nonlinear programming and minimax problems at an isolated local solution is discussed. The correct identification of active constraints can improve the local convergence behavior of algorithms and considerably simplify algorithms for inequality constrained problems, so it is a useful adjunct to nonlinear optimization algorithms. Facchinei et al. [F. Facchinei, A. Fischer, C. Kanzow, On the accurate identification of active constraints, SIAM J. Optim. 9 (1998) 14–32] introduced an effective technique which can identify the active set in a neighborhood of a solution for nonlinear programming. In this paper, we first improve this conclusion to be more suitable for infeasible algorithms such as the strongly sub-feasible direction method and the penalty function method. Then, we present the identification technique of active constraints for constrained minimax problems without strict complementarity and linear independence. Some numerical results illustrating the identification technique are reported.