最优化中的几个问题解的等价性

考虑如下非线性规划问题：众所周知，问题（NP）的解法主要有三类：1．直接处理约束，2．将约束最优化问题化为 无约束最优化问题来处理，3．将（NP）化为简单的约束最优化问题如线性规划或二次规划等来处理，而将约束最优化问题化为无约束最优化问题的主要手段是利用如下的Lagrange函数：L（X，X，X）一八X）＋（X，g（X》十（X，h（X》（1．I）定义1．1称点卜”，V”撤足互补性条件，如果对”（X）一ojE【I：c］（亚．2）根据Lagrange函数（1．1）定义如下问题：（SPP）：求点k”，u”，v」6H””，m二。；＋c，使b“，u“，v」…     

Some minimax theorems in vector-valued functions

In this paper, we generalize the well-known notions of minimax, maximin, and saddle point to vector-valued functions. Conditions for a vector-valued function to have a generalized saddle point are given. An example is used to illustrate the generalized concepts of minimax, maximin, and saddle point.     

On a pair of nonlinear mixed integer programming problems

We consider maximin and minimax nonlinear mixed integer programming problems which are nonsymmetric in duality sense. Under weaker (pseudo-convex/pseudo-concave) assumptions, we show that the supremum infimum of the maximin problem is greater than or equal to the infimum supremum of the minimax problem. As a particular case, this result reduces to the weak duality theorem for minimax and symmetric dual nonlinear mixed integer programming problems. Further, this is used to generalize available results on minimax and symmetric duality in nonlinear mixed integer programming.     

Minimax designs for linear regression models with bias in a reproducing kernel Hilbert space in a discrete set

Consider the design problem for estimation and extrapolation in approximately linear regression models with possible misspecification. The design space is a discrete set consisting of finitely many points, and the model bias comes from a reproducing kernel Hilbert space. Two different design criteria are proposed by applying the minimax approach for estimating the parameters of the regression response and extrapolating the regression response to points outside of the design space. A simulated annealing algorithm is applied to construct the minimax designs.These minimax designs are compared with the classical D-optimal designs and all-bias extrapolation designs. Numerical results indicate that the simulated annealing algorithm is feasible and the minimax designs are robust against bias caused by model misspecification.     

Some interesting properties of maximin strategies

Since the seminal paper of Nash (1950) game theoretic literature has focused mostly on equilibrium and not on maximin (minimax) strategies. We study the properties of these strategies in non-zero-sum strategic games that possess (completely) mixed Nash equilibria. We find that under certain conditions maximin strategies have several interesting properties, some of which extend beyond 2-person strategic games. In particular, for n-person games we specify necessary and sufficient conditions for maximin strategies to yield the same expected payoffs as Nash equilibrium strategies. We also show how maximin strategies may facilitate payoff comparison across Nash equilibria as well as refine some Nash equilibrium strategies.     

One-dimensional nested maximin designs

The design of computer experiments is an important step in black-box evaluation and optimization processes. When dealing with multiple black-box functions the need often arises to construct designs for all black boxes jointly, instead of individually. These so-called nested designs are particularly useful as training and test sets for fitting and validating metamodels, respectively. Furthermore, nested designs can be used to deal with linking parameters and sequential evaluations. In this paper, we introduce one-dimensional nested maximin designs. We show how to nest two designs optimally and develop a heuristic to nest three and four designs. These nested maximin designs can be downloaded from the website . Furthermore, it is proven that the loss in space-fillingness, with respect to traditional maximin designs, is at most 14.64 and 19.21%, when nesting two and three designs, respectively.     

Efficient Location for a Semi-Obnoxious Facility

This paper deals with a location model for the placement of a semi-obnoxious facility in a continuous plane with the twin objectives of maximizing the distance to the nearest inhabitant and minimizing the sum of distances to all the users (or the distance to the farthest user) in a unified manner. For special cases, this formulation includes (1) elliptic maximin and rectangular minisum criteria problem, and (2) rectangular maximin and minimax criteria problem. Polynomial-time algorithms for finding the efficient set and the tradeoff curve are presented.     

Local Polynomial Regression: Optimal Kernels and Asymptotic Minimax Efficiency

We consider local polynomial fitting for estimating a regression function and its derivatives nonparametrically. This method possesses many nice features, among which automatic adaptation to the boundary and adaptation to various designs. A first contribution of this paper is the derivation of an optimal kernel for local polynomial regression, revealing that there is a universal optimal weighting scheme. Fan (1993, Ann. Statist., 21, 196-216) showed that the univariate local linear regression estimator is the best linear smoother, meaning that it attains the asymptotic linear minimax risk. Moreover, this smoother has high minimax risk. We show that this property also holds for the multivariate local linear regression estimator. In the univariate case we investigate minimax efficiency of local polynomial regression estimators, and find that the asymptotic minimax efficiency for commonly-used orders of fit is 100% among the class of all linear smoothers. Further, we quantify the loss in efficiency when going beyond this class.     

投影均匀分片拉丁超立方体设计

空间填充设计是有效的计算机试验设计,比如均匀设计、最大最小距离拉丁超立方体设计等.虽然这些设计在整个试验空间中有较好的均匀性,但其低维投影均匀性可能并不理想.对于因子是定量的计算机试验,已有文献构造了诸如最大投影设计、均匀投影设计等相适应的设计;而对于同时含有定性因子和定量因子的计算机试验,尚未有投影均匀设计的相关文献.文章提出了综合投影均匀准则,利用门限接受算法构造了投影均匀的分片拉丁超立方体设计.在新构造设计中,整体设计与每一片设计均具有良好的投影均匀性.模拟结果显示,与随机分片拉丁超立方体设计相比,利用新构造设计进行试验而拟合的高斯过程模型具有更小的均方根预测误差.     

Asymptotically Efficient Sequential Kernel Estimates of the Drift Coefficient in Ergodic Diffusion Processes

A sequential asymptotically efficient procedure is constructed for estimating the drift coefficient at a given state point in ergodic diffusion processes. Sequential kernel estimators are used. The optimal convergence rate with the sharp constant is given for a local minimax risk.     

Interior penalty functions and duality in linear programming

Logarithmic additive terms of barrier type with a penalty parameter are included in the Lagrange function of a linear programming problem. As a result, the problem of searching for saddle points of the modified Lagrangian becomes unconstrained (the saddle point is sought with respect to the whole space of primal and dual variables). Theorems on the asymptotic convergence to the desired solution and analogs of the duality theorems for the arising optimization minimax and maximin problems are formulated.     

Unification of differential games,generalized solutions of the Hamilton-Jacobi equations,and a stochastic guide

On the basis of a minimax-maximin differential game of minimal deviation of a moving object from a given target on a given time interval, we discuss the relationship between the unified form of the game and the interpretation of this game in terms of Subbotin’s generalized minimax solution of Hamilton-Jacobi equations. In addition, we consider the relationship between the chosen formalization of a differential game and investigations of this game on the basis of solutions of parabolic equations degenerating into Hamilton-Jacobi equations as the diffusion term tends to zero. The related generation of minimax and maximin controls with a stochastic guide is described. We analyze the similarity between the unified form of a differential game and the concept of differential game suggested by Pontryagin.     

Some topological minimax theorems

Two topological variants of the minimax theorem are proved with no restrictions on one of the spaces except for those related to the function under consideration. The conditions concerning the behavior of the function deal only with the interval between the maximin and minimax. As corollaries, we obtain the well-known theorems of Sion and Hoang-Tui on quasiconvex-quasiconcave semicontinuous functions. The scheme of arguments goes back to the Hahn-Banach theorem and the separating hyperplane theorem. It is shown how this scheme can be explicitly realized in the proof of the Hahn-Banach theorem. Translated fromMaternaticheskie Zametki, Vol. 67, No. 1, pp. 141–149, January, 2000.     

Asymptotic minimax estimation in semiparametric models

Summary We give several conditions on the estimator of efficient score function for estimating the parametric component of semiparametric models. A semiparametric version of the one-step MLE using an estimator of efficient score function which fulfills the conditions is shown to converge to the normal distribution with minimum variance locally uniformly over a fairly large neighborhood around the assumed semiparametric model. Consequently, it is shown to be asymptotically minimax with bounded subconvex loss functions. A few examples are also considered.     

The facility location problem when the underlying distribution is either dominated or dominating

The theory of choices involving Risk is used to solve some classes of the Today maximin and minimax one facility location problems. Solutions are shown when the underlying demand distribution is symmetric and unimodal.     

Two-dimensional minimax Latin hypercube designs

We investigate minimax Latin hypercube designs in two dimensions for several distance measures. For the ?^∞-distance we are able to construct minimax Latin hypercube designs of n points, and to determine the minimal covering radius, for all n. For the ?¹-distance we have a lower bound for the covering radius, and a construction of minimax Latin hypercube designs for (infinitely) many values of n. We conjecture that the obtained lower bound is attained, except for a few small (known) values of n. For the ?²-distance we have generated minimax solutions up to n=27 by an exhaustive search method. The latter Latin hypercube designs are included in the website www.spacefillingdesigns.nl.     

Multicriteria Semi-Obnoxious Network Location Problems (MSNLP) with Sum and Center Objectives

Locating a facility is often modeled as either the maxisum or the minisum problem, reflecting whether the facility is undesirable (obnoxious) or desirable. But many facilities are both desirable and undesirable at the same time, e.g., an airport. This can be modeled as a multicriteria network location problem, where some of the sum-objectives are maximized (push effect) and some of the sum-objectives are minimized (pull effect).We present a polynomial time algorithm for this model along with some basic theoretical results, and generalize the results also to incorporate maximin and minimax objectives. In fact, the method works for any piecewise linear objective functions. Finally, we present some computational results.     

A multistage pursuit-evasion game that admits a gaussian random process as a maximin control policy

In this paper we consider the existence and structure of both minimax and maximin policies for the special class of LQG pursuit-evasion games which is characterized by (i) a blind evader; and (ii) a pursuer who can make use of noise corrupted state measurements. The particular class of games which we consider has been studied previously by other investigators who have shown that pure strategies exist for both players. The major contribution of our paper is the delineation of the existence and structure of a mixed strategy for the evader in this class of games. This new maximin strategy is defined by a gaussian measure, which can be determined explicitly by the method of least favorable prior distributions. We show that the validity of the pure solutions determined previously is limited by the duration of the game, due to the existence of a ‘pure solution conjugate point’; further, we prove that our new strategies are valid solutions which extend the possible duration of the game beyond the limit imposed by the pure solution conjugate point. We believe that our paper constitutes the first report on the existence of a mixed strategy for an LQG game, and the first report on the role conjugate points play in the transition between pure strategies and mixed strategies.