Title and Abstract

Plenary Talk
**************

AHMED, N.U.
Dynamics of Computer Communication Networks and Optimal Control

Abstract : For details, please click here.


CHAO, Xiuli
Optimal Control Policies for Inventory Systems Subjecting to Supplier Incentives

Abstract : Classical inventory systems and their optimal control policies are based on the traditional ordering cost structure of setup cost and linear purchasing cost. However, such cost structure is no longer valid as shown in the energy crisis in the US in the past several years, in which the energy supplier offers incentives for users to not over use its capacity.  In this talk we present several mathematical models to study such systems and derive their optimal control policies.  Our results extend the classical results in inventory theory.


CHENG, Daizhan
Quadratic Stability and Stabilization of Switching Linear Systems

Abstract : For details, please click here.


DENG, Xiaotie
Algorithmic Issues in Economics and Finance

Abstract : Computation has without doubt long been an important aspect in economics and finance.  The approach of algorithms and complexity developed in computer science, however, provide a distinguished viewpoint in the understanding of the principles in economics and finance.  In this talk, we discuss some of the recent development in this area and propose new directions and open problems.


DING, Jiu
The Maximum Entropy Method and Its Applications to Numerical Dynamical Systems

Abstract : The maximum entropy method (maxent) is widely used in the context of the moment problem which appears naturally in many branches of physics and engineering; it is used to numerically recover the density with least bias from finitely many known moments.  In this talk we introduce the basic idea behind this method and discuss some issues on the convergence.  Finally, we apply this method to solving operator equations for densities, in particular the fixed density problem for Markov operators in stochastic analysis and Frobenius-Perron operators in ergodic theory.


EKELAND, Ivar
Duality in Optimal Transportation and the Structure of Cities

Abstract : The mathematical theory of optimal transportation consists in seeking a measure-preserving map which maximizes some integral criterion.  This theory has found a very nice dual formulation in recent years (Kantorovitch, Brenier, Gangbo, McCann).  I wish to describe new results of Guillaume Carlier and myself. Building upon a one-dimensional model of Robert E. Lucas Jr and Esteban Rossi-Hansberg, we are able to formulate in full generality an equilibrium model for the structure of cities (i.e. the repartition of the ground between residential and business districts) and to solve it by a fixed-point theorem.


FUKUSHIMA, Masao
Hybrid Simulated Annealing and Direct Search Method for Nonlinear Global Optimization

Authors: Abdel-Rahman Hedar and Masao Fukushima (Kyoto University)

Abstract : In this paper, we give a new approach of hybrid direct search methods with meta-heuristics of simulated annealing for finding a global minimum of a nonlinear function with continuous variables.  First, we suggest a Simple Direct Search (SDS) method, which comes from some ideas of other well known direct search methods.  Since our goal is to find global minima and the SDS method is still a local search method, we hybridize it with the standard simulated annealing to design a new method, called the Simplex Simulated Annealing (SSA) method, which is expected to have some ability to look for a global minimum.  To obtain faster convergence, we first accelerate the cooling schedule in SSA, and in the final stage, we apply Kelley's modification of the Nelder-Mead method on the best solutions found by the accelerated SSA method to improve the final results.  We refer to this last method as the Direct Search Simulated Annealing (DSSA) method.  The performance of SSA and DSSA is reported through extensive numerical experiments on some well known functions.  Comparing their performance with that of other meta-heuristics shows that SSA and DSSA are promising in practice.  Especially, DSSA is shown to be very efficient and robust.


GAO, David
Complete Solutions and Triality Theory for Global Optimization with Applications

Abstract : Duality is a fundamental concept that underlies almost all natural phenomena.  Triality, a newly discovered theory, reveals the intrinsic pattern of dualities in complex systems.  Based on the powerful canonical dual transformation method developed recently [1], the speaker will show that many very difficult nonsmooth/ nonconvex optimization problems in n-dimensional space can be converted into certain smooth, picewise convex dual problems in lower dimensional space.  Therefore, complete solutions, including all extremum points (both minimizers and maximizers) and all saddle points, are obtained for a large class of nonconvex optimization problems.  The extremality conditions for all these solutions are controlled by the triality theory.  Applications are illustrated by examples from nonsmooth d.c. programming, chaotic dynamical systems and phase transitions in physics and material science.


KELLEY, C.T.
The Implicit Filtering Method for Noisy Optimization Problems

Abstract : Implicit filtering is a sampling method that uses difference gradients in a projected BFGS iterations.  The size of the difference increment is reduced as the optimization progresses.  The objective of this is to avoid entrapment in local minima of functions which have many local minima.  In this talk we will discuss work on applications to groundwater remediation and some recent theoretical results.


KOJIMA, Masakazu
A General Framework for Convex Relaxation of Polynomial Optimization Problems over Cones

Abstract : The class of POPs (polynomial optimization problems) over cones covers a wide range of optimization problems such as 0-1 integer linear and quadratic programs, nonconvex quadratic programs and bilinear matrix inequalities.  This paper presents a new framework for convex relaxation of POPs over cones in terms of linear optimization problems over cones.  It provides a unified treatment of many existing convex relaxation methods based on the lift-and-project linear programming procedure, the reformulation-linearization technique and the semidefinite programming relaxation for a variety of problems.  It also extends the theory of convex relaxation methods, and thereby brings flexibility and richness in practical use of the theory.


LU, Qiang
An Approach to Nonlinear Robust Control and Its Application to Power System

Abstract : For details, please click here.


LUO, Zhiquan
Robust Optimization Techniques for Multi-Antenna Beamforming

Abstract : In this talk, I will describe some recent advances in the robust optimization techniques for multi-antenna beamforming applications.  These new robust beamforming methods can operate very effectively in the presence of unknown arbitrary-type mismatches of the desired signal array response, and they significantly outperform the existing beamforming methods.


MOORE, John B.
Pose Estimation of known 3D Objects and Camera Focal Length from 2D Images

Abstract : A quadratically convergent least-squares approach is enhanced for pose estimation of a known cluster of points in 3D, the object, which is observed by a single calibrated camera as a 2D cluster of points, the image.  The optimization-on-a-manifold approach discussed also applies to the mathematically equivalent task of the calibration of the extrinsic parameters of a monocular vision system.  As with all such pose estimation methods, the observed object must consist of at least three 3D points, and four generic points for a unique solution.  Analytical solutions, or solutions derived from a singular value decomposition, are used for the noise free case and for initializations in low noise.

The algorithms in their simplest form, when focal length is known, are an optimization over three parameters, and typically converge to a minima with reasonable accuracy  in two iterations.  The more generic image points, the greater is the probability that this is a global minimum.  A line search feature, including in a random direction ensures that with more iterations, the probability of 'global' convergence increases.  The global minimum are known to be robust to model uncertainty and-or measurement noise, with robustness properties that can be tuned via weighting matrix selection.

Generalization of the approach to deal with focal length uncertainty is included.  Generalization to stereo and multiple camera pose estimation tasks is straightforward, although for these camera calibrations are required for these.  It is quite realistic to apply the algorithms for on-line video processing.


MOTREANU, Dumitru
Nonsmooth Variational Methods with Applications to Nonlinear Problems in Mechanics

Abstract : For details, please click here.


PARDALOS, Panos M.
Multi-Variable Partition Approach for Optimization Problems

Abstract : In this talk we present a Multi-Variable Partition (MVP) approach for dealing with computationally hard optimization problems.  The MVP approach partitions all variables appearing in an optimization problem into several groups, and regards each group as a macro-variable.  We will consider applications of this approach for solving molecular conformation problems and the spherical code problem.

This is joint work with Prof. Hong-Xuan Huang from Tsinghua University, Beijing.

References

P.M. Pardalos, D. Shalloway and G. Xue,
Global Minimization of Nonconvex Energy Functions: Molecular Conformation and Protein Folding,
DIMACS Series Vol. 23, American Mathematical Society, (1996).

H. X. Huang, P.M. Pardalos, and Z.J. Shen,
A point balance algorithm for the spherical code problem,
Journal of Global Optimization Vol. 19, No. 4 (2001), pp. 329-344.

H. X. Huang, P.M. Pardalos, and Z.J. Shen,
Equivalent formulations and necessary optimality conditions for the Lenard-Jones problem,
Journal of Global Optimization, Vol. 22, (2002), pp. 97-118.

H. X. Huang and P.M. Pardalos,
Multivariate Partition Approach for Optimization Problems,
Cybernetics and System Analysis (2002).


QI, Liqun
Multivariate Polynomial Minimization and Its Application in Signal Processing

Abstract : For details, please click here.


RUBINOV, Alex M.
Lagrange-type Functions in Constrained Optimization

Abstract : For details, please click here.


SUN, Wenyu
On Trust Region Method and Line Search Approach

Abstract : Trust region method is an important class of methods for optimization.  This paper studies the relationship between the trust region method and the line search approach, exposes that the trust region method is, in fact, a generalization and development of Armijo line search rule, and consequently it is descent sufficiently and convergent globally.  Since trust region method employs the constrained quadratic function model and updating rule of the trust region radius, it is more robust and efficient for optimization problems than line seach approach.  From the idea, as an extension, the author discusses further the nonmonotone trust region method and the second order trust region method.


TEO, Kok Lay
Optimal Control Problems Subject to Multiple Characteristic Time Constraints

Abstract : For details, please click here.


THERA, Michel
Variational Composition of a Monotone Operator with a Linear Mapping

Abstract : This lecture concerns a recent joint work with T. Pennanen and J.-P. Revalski.  We will propose a regularized notion of composition of a monotone operator with a linear mapping.  This new concept, called variational composition, can be shown to be maximal monotone in many cases where the usual composition is not.  The two notions coincide, however, whenever the latter is maximal monotone.  The utility of the variational composition is shown by several applications, including subdifferential calculus, theory of measurable multifunctions and elliptic PDEs with singular coefficients.


WANG, Shouyang
Multiperiod Portfolio Selection Based on a Minimax Rule

Abstract : For details, please click here.


WU, Soon-yi
Analytic Center Cutting Plane Methods for Solving Semi-Infinite Variational Inequality Problems

Abstract : For details, please click here.


XU, Chengxian
Models of Optimal Hedging Ratio with Transaction Costs

Abstract : In making a decision on hedging, it is important to know the optimal hedging ratio to minimize the price risk of investments.  This talk presents the models that determine the optimal hedging ratio under the consideration of transaction costs.  These include the models of minimizing the variance of hedging strategy, the models of optimal hedging ratio with goals in either returns or risks and the models of multi-period hedging strategy.  Unlike the existing models, all the models presented in the talk consider the transaction costs.


YANG, Xiaoqi
Penalty Type Methods for Mathematical Programs with Complementarity Constraints

Abstract : For details, please click here.


ZHANG, Shuzhong
A New Self-Dual Embedding Method for Convex Programming

Abstract : In this paper we introduce a conic optimization formulation for inequality-constrained convex programming, and propose a self-dual embedding model for solving the resulting conic optimization problem.  The primal and dual cones in this formulation are characterized by the original constraint functions and their corresponding conjugate functions respectively.  Hence they are completely symmetric.  This allows for a standard primal-dual path following approach for solving the embedded problem.  Moreover, there are two immediate logarithmic barrier functions for the primal and dual cones.  We show that these two logarithmic barrier functions are conjugate to each other.  The explicit form of the conjugate functions are in fact not required to be known in the algorithm.  An advantage of the new approach is that there is no need to assume an initial feasible solution to start with.  To guarantee the polynomiality of the path-following procedure, we may apply the self-concordant barrier theory of Nesterov and Nemirovski.  For this purpose, as one application, we prove that the barrier functions constructed this way are indeed self-concordant when the original constraint functions are convex and quadratic.


Other Talk
*************




BAI, Min-ru
On a class of set-valued variational inclusions in Banach spaces

Abstract : For details, please click here.


CHAO, Xiuli
Optimal resource allocation in multisite service systems with intersite customer flows

Abstract : We study the optimal resource allocation problem in multisite service systems with intersite customer flows.  The problem is studied using Stackelberg game.  The optimal resource allocation as well as the corresponding customer reaction rules are characterized.  The optimal solution is in sharp constrast with the commonly used proportional rules, and it has the structure of "one big, and many small".  This is a joint work with Liming Liu and Shaohui Zheng of HK University of Science and Technology.


CHEN, Guang-ya
Vector Network Equilibrium Problems and Nonlinear Scalarization Methods

Abstract : The earliest network equilibrium model was proposed by Wardrop (1952) for a transportation network.  Since then, many other equilibrium models have also been proposed in the economic literature (See Narguney, 1993).  In this talk, equilibrium models based on multicriteria consideration or vector-valued cost functions have been proposed.  Some nonlinear scalarization methods have been introduced.  Other results that consider multicriteria equilibrium models can be found in Dial (1996), Laurent (1993), Chen and Yen (1993) and X.Q. Yang and C.J. Goh (1997) and so on.


DENG, Sien
Convex Analysis of Vector Optimization Problems

Abstract : Convex analysis is a powerful tool in continuous optimization. Here it is employed to study vector optimization problems. The aim of this talk is to illustrate how the fundamental results in convex analysis can be effectively used to investigate sensitivity in convex vector optimization problems. Although solution sets involved in such problems are usually non-convex, they are highly structured.  By examining such structures carefully, we are able to obtain a number of positive results and sharpen some existing results in the literature. We then apply these results to the problem of parametric vector optimization.


FRANCA, Felipe M.G.
A Multistart Approach to Near-optimal Concurrency Dynamics in Neighborhood-constrained Systems

Abstract : For details, please click here.


FENG, Guosheng

Combination of trust region method and simplicial decomposition for convex constrained nonlinear problems

Abstract : For details, please click here.


GASIMOV, Rafail N.
Strictly increasing positively homogeneous functions with applications to exact penalization

Abstract : For details, please click here.


GE, Rendong
An Modified ABS Algorithm for Solving Singular Nonlinear Systems with Rank One Defect (1)

Abstract : For details, please click here.

An Type of ABS Algorithm for Solving Singular Nonlinear Systems with Rank Defects (2)

Abstract : For details, please click here.


HILL, David John
Optimisation for Control of Complex Systems

Abstract : The subject of control is evolving to handle greater levels of system complexity, namely large scale (or dimension), strong nonlinearity, uncertainty and heterogeneous nature (mixed models, control requirements).  Applications include many kinds of processes in electrical engineering, manufacturing, biomedical engineering and finance.  Models in general are of the hybrid kind, i.e. a mixture of differential, algebraic and switching equations with different versions in different spatial, parametric and state domains.  A framework for the design of truly global controllers, which act in a coordinated way across all the regions of operation in terms of both scale, i.e. all states, and in the large with respect to disturbances and model variations, requires sophisticated optimisation merged with ideas from modern control.  This talk presents progress towards development of such a framework, which is called global optimal control.

The scheme that will be presented for global control is of hierarchical form: 1) use of hybrid models; 2) local optimal control elements with tunable parameters; 3) computation of  bifurcation boundaries; 4) the  global control structure; 5) optimal coordination.  This combination achieves a high-level version of distributed adaptive optimal control which "swarms" around the complex system attacking problems as they arise.  The common learning control paradigms, including adaptive and fuzzy-neural control are special cases which will not work globally except in special circumstances.  Optimal vs suboptimal, but practically feasible, techniques will be discussed.


HOU, Shui-hung
On the Existence of Solutions for a Class of Nonlinear Inclusions

Abstract : For details, please click here.


HU, Cheng-feng
Solving Fuzzy Variational Inequalities

Abstract : This work studies variational inequalities for fuzzy mappings over a fuzzy domain.  It is shown that such problems can be reduced to bilevel programming problems.  A penalty function algorithm is introduced with a convergence proof, and numerical examples are included to illustrate the solution procedure.


HUANG, Xue-xiang
Partial Augmented Lagrangian Method and Mathematical Programs with Complementarity Constraints

Abstract : For details, please click here.


ITO, Satoshi
Tracking Control of General Nonlinear Systems by a Direct Gradient Descent Method

Abstract : For details, please click here.


JIAN, Jin-bao
A Superlinearly and Quadratically Convergent SSLE Algorithm for Optimization Problems without Strict Complementarity

Abstract : For details, please click here.


KIM, Sunyoung
SDP and SOCP relaxations of some class of quadratic optimization problems

Sunyoung Kim
Department of Mathematics, Ewha Women's University
11-1 Dahyun-dong, Sudaemoon-gu, Seoul 120-750 Korea
email: {\it skim@ewha.ac.kr}

and

Masakazu Kojima
Department of Mathematical and Computing Sciences
Tokyo Institute of Technology
2-12-1 Oh-Okayama, Meguro-ku, Tokyo 152-8552 Japan
email:  kojima@is.titech.ac.jp

Abstract : SDP (Semidefinite Programming) relaxation is known to provide effective bounds for objective values of nonconex QOPs (Quadratic Optimization Problems), while lift-and-project LP (Linear Programming) gives inferior bounds with faster computing time.  SOCP (Second Order Cone Programming) relaxation falls between the effectiveness of SDP relaxation and efficiency of lift-and-project LP relaxation.  Two types of SOCP relaxation are discussed to illustrate the effectiveness and efficiency of SOCP relaxation.  We also present a class of QOPs whose exact optimal values can be obtained via SDP and SOCP relaxations.


LI, Jian-ling
An Interior-point Continuation Method for Monotone Variational Inequalities with Box Constraint

Abstract : For details, please click here.


LI, Leong-kwan
Learning Sunspot Series Dynamics by Recurrent Neural Networks

Abstract : For details, please click here.


LI, Sheng-jie
A Solution Method for Generalized Semi-Infinite Programming

Absract : For details, please click here.


LI, Yihua
Jobshop Scheduling with Variable Duration and Multiple Resources

Abstract : For details, please click here.


LI, Zhong-fei
A Closed-Form Solution to a Dynamic Portfolio Optimization Problem

Abstract : For details, please click here.


LIU, Guoqing
Nonlinear Retrieval of Wind Field via Optimal Technique

Abstract : For details, please click here.


LIU, Hailong
Optimal Control of Porfolio Liquidation

Abstract : For details, please click here.


LIU, Jun
Optimal Sensor Locations for Ill-posed Inverse Problems

Abstract : Inverse problems, which involve the determination of some unknown physical quantities from indirect experimental data, arise in very diverse fields of science and engineering, such as geophysics, heat transfer, medical diagnostics, meteorology, mechanics {\it etc}. One main difficulty in solving such inverse problems is that they are ill-posed. In a physical experiment, the data are obtained from measurements of sensors.  Different sensor locations could make the inverse problem less or more ill-posed.  Our aims is to determined the optimal sensor locations so that the inverse problem is as less ill-posed as possible.

Mathematically, given sampling points $s_{i}$, for $i=3D 1,\dots,m$, a Fredholm integral equation of the first kind can be discretized into a semi-discrete form  $g(s_{i})=3D\int_{0}^{1} k(s_{i},t)f(t)dt$ by using a collocation method=, and into a fully discrete form $g(s_{i})=3D\int_{0}^{1}k(s_{i},t)f_{n}(t)dt$ by restricting the unknown function $f_{n}(t)$ within an  $n$-dimensional subspace.  Our optimal design problem is to determine a distribution of the discrete points $s_{i},~i=3D1,\dots,m$ maximizing the $n$th singular value $\la_{n}$ of the semi-discrete or the fully discrete operator $A(s)$ where the number $n$ plays the role of the regularization parameter.  The optimal sensor locations problem for various inverse problems, such as the inverse scattering problem, is  studied numerically.Optimal Sensor Locations for Ill-posed Inverse Problems


LIU, Yanqun
An Efficient Dual Parametrization Method for Quadratic Semi-infinite Programming Problems

Abstract : For details, please click here.


PU, Dingguo
A New QP-Free Infeasible Method for Nonlinear Inequality Constrained Optimization Problems

Abstract : In this paper, a new QP-free infeasible method is proposed for minimizing a smooth function subject to smooth inequality constraints.  This iterative method is base on the solution of nonsmooth equations wich  are obtained by the multiplier and some NCP functions for the KKT first-order optimality conditions.  Locally, each iteration of this method can be viewed as a perturbation of a Newton or quasi-Newton iteration on both the primal and dual variables for the solution of the KKT optimality conditions.  This method does not request feasibility for the iterates.  In particular, we define a merit function which is related to the objective function, the multiplier function and the NCP function.  Then we make the merit function decreasing at each iteration.  This method is  implementable and globally convergent without assuming the strict complementarity condition, isolatedness of the accumulation point and linear independence of the gradients of active constrained functions at the solution.  We also prove that the method has superlinear convergence rate under some mild conditions.  Some preliminary numerical results indicate that this new QP-free infeasible method is quite promising.


RATIU, Tudor S.
Controllability of Reduced Systems

Abstract : For details, please click here.


RUBINOV, Alex M.
Cluster function and its minimization

Authors : A.M. Bagirov and A. M. Rubinov

Abstract : The search for clusters of a finite set can be reduced to the minimization of a certain saw-tooth function, which is called cluster function.  We discuss some properties of the cluster function.  It follows from these properties that the derivative-free discrete gradient method can be used for the minimization of the cluster function.  We give a short description of this method and discuss results of numerical experiments .


SCHULZ, Lech
On the fuzzy set optimization of complex systems

Abstract : For details, please click here.


SHI, Jianming
Quasiconjuagte method and its applicatoin to the sum of ratios problem

Abstract : In this paper, we present a method for a reverse convex programming problem.  We suppose that the feasible region of our promblem is a d.c set, that is, a difference of two convex sets $D$ and  $C$.  We use  a sequence of  polytopes $P_k$ to approximate the set $C$ and relax the problem to one which can be solved by quasiconjuagte methods.  The algorithm generates a sequence of solution points of the relaxed problems.  Each accumulation point of the generated points is an optimal solution of our problem.  As an application, we will discuss the sum of ratios problem.


SHI, Zhenjun
On M-Step Memory Gradient Method with Wolfe's Line Search

Abstract : For details, please click here.


TONG, Xiaojiao
On the Convergence of a Trust Region Method for Solving Constrained Nonlinear Equations

Abstract : For details, please click here.


VIKAS, B.L.
Development of Software Package for Project Management Using CPM and Pert with Simulation

Abstract : The complexity of the present day management problem and the business competition has added to the already existing pressure on the brains of decisions makers.  In a large and complex projects involving a number of interrelated activities, requiring a number of men machines and materials.  It is not possible for the management to make and execute optimum schedule just by invitation based on the organizational capabilities and work experience.

Management are thus always on the look out for the methods and technologies, which may help in planning, scheduling the project.  A project may be defined as a combination of interrelated activities, which must be executed in a certain order before the entire task can be completed.

The aim of the planning is to develop a sequence of activities of the project, so that the project completion time and cost are properly balanced and that the excessive demand of key resources are avoided.  To meet the object of systematic planning, the management has evolved a number of techniques applying network strategy.

Program evaluation review technique [PERT] and the critical path method [CPM] are two of the many network techniques which have been widely used for planning, scheduling and controlling the large and complex project] are two of the many network techniques which have been widely used for planning, scheduling and controlling the large and complex project.  The technique that we have followed for drawing the network is Activity on Node diagram [AON] which is more simplified than the older Activity on Arrow diagram[AOA].  The project manager for managing a project has to manually draw the network and apply network strategy to arrive  at the critical and minimum completion time.

To make his job easier and accurate, the purpose of devising software becomes apparent.  For this purpose, visual basic-6 and MS-ACCESS have been used. Also to determine the optimum project schedule, crashing program is also included as cost of completion of the project is also important in completing the projects.

As the day to day activities are uncertain, one has to complete for the uncertainties that may take place in the industry.  For which we have added the SIMULATION  PACKAGE which makes use of Monte Carlo technique.  This makes the package more complete providing the project manager a foresight of what might happen and let him take decision  accordingly to counter or contain risks.

The specialty of our project is that we have used AON diagram technique to determine/calculate Total float, Free float , and Independent float, which does not exist in other similar packages.


WAN, Zhong
Convergence of An Inexact Smoothing Continuation Method for a class of MPEC

Abstract : For details, please click here.


WANG, Jinliang

Abstract : For details, please click here.


WANG, Yiju
A Nonsmooth L-M Method for the Solution of the Generalized Complementarity Problem

Abstract : For details, please click here.


WEI, Wei
On existence of solution for a class of strongly nonlinear integro-differential equation

Abstract : For details, please click here.


YAN, Houmin
A Periodic Review Inventory Model with Three Delivery Modes and Forecast

Guillermo Gallego
Dept. of Industrial Engineering and Operations Research
The ColumbiaUniversity in the City of New York

Ying Huang
T.J.WatsonResearchCenter
Yorktown Heights, N.Y.10598

Suresh P. Sethi
School of Management, the University of Texas at Dallas
Richardson, TX75083-0688

Houmin Yan
Dept of Systems Engineering and Engineering Management
TheChineseUniversity of Hong Kong
Shatin, N.T., Hong Kong

Hanqin Zhang
Institute of Applied Mathematics
Academy of Mathematics and Systems Sciences
Academia Sinica, Beijing, 100080, China

Abstract : This paper is concerned with a periodic review inventory system with three delivery modes and demandforecast updates.  At the beginning of each period, on-hand inventory and demand information are updated.  At the same time, decisions on how much to order using three delivery modes are made.  It is shown that there exists an optimal Markov policy and that it is a modified base-stock policy.


YANG, Xinmin
Generalized Invexity and Generalized Invariant Monotonicity

Abstract : For details, please click here.


YIU, Cedric
Near-field broadband beamformer optimal design

Abstract : For details, please click here.


YU, Bo
The Aggregate Homotopy Method for Unconstrained Sequential Min-Max Problem

Abstract : For details, please click here.


YUE, Wuyi
Optimal Replacement of a System in a Semi-Markov Environment

Abstract : For details, please click here.


ZASLAVSKI, Alexander  J.
Well-Posedness and Porosity in Optimal Control Without Convexity Assumptions

Abstract : For details, please click here.

On a Generic Existence Result in Optimization

Abstract : For details, please click here.


ZHANG, Roxin
Multistage Bilevel Problems

Abstract : In this presentation, we discuss the concepts of bilevel and n-level mathematical programming problems and give some ideas in formulating multistage bilevel problems.  The current techniques for obtaining the optimality conditions of single stage bilevel problems are surveyed and compared.  And finally we attempt to apply some of these techniques to multistage problems.


ZHOU, Houchun
Duality without Constraint Qualification for Minimax Fractional Programming

Author : Houchun Zhou* and Wenyu Sun

Abstract : Without the need of constraint qualification, we establish the necessary and sufficient optimality conditions for generalized fractional programming involving a compact set.  Using these  optimality conditions, we construct a parametric dual model and a parameter-free mixed dual model, this mixed dual model unify two dual parameter-free models constructed by Lai and Liu et al.  Several duality theorems are established.


ZOU, Zhiqiang
An approximation approach to optimal control problems

Abstract : For details, please click here.


Updated on 7 August 2002