Showing papers in "Mathematical Methods of Operations Research in 2005"

Benchmark and mean-variance problems for insurers

Abstract: We consider the classical Cramer-Lundberg model with dynamic proportional reinsurance and solve the problem of finding the optimal reinsurance strategy which minimizes the expected quadratic distance of the risk reserve to a given benchmark. This result is extended to a mean-variance problem.

Lipschitz Continuity of Value Functions in Markovian Decision Processes

Abstract: We present tools and guidelines for investigating Lipschitz continuity of the value functions in MDP’s, using the Hausdorff metric and the Kantorovich metric for measuring the influence of the constraint set and the transition law, respectively. The methods are explained by examples. Additional topics include an application to the the discretization algorithm of Bertsekas (1975).

Worst-Case Scenario Portfolio Optimization: a New Stochastic Control Approach

Abstract: We consider the determination of portfolio processes yielding the highest worst-case bound for the expected utility from final wealth if the stock price may have uncertain (down) jumps. The optimal portfolios are derived as solutions of non-linear differential equations which itself are consequences of a Bellman principle for worst-case bounds. A particular application of our setting is to model crash scenarios where both the number and the height of the crash are uncertain but bounded. Also the situation of changing market coefficients after a possible crash is analyzed.

Simulation of typical Cox–Voronoi cells with a special regard to implementation tests

Abstract: We consider stationary Poisson line processes in the Euclidean plane and analyze properties of Voronoi tessellations induced by Poisson point processes on these lines. In particular, we describe and test an algorithm for the simulation of typical cells of this class of Cox–Voronoi tessellations. Using random testing, we validate our algorithm by comparing theoretical values of functionals of the zero cell to simulated values obtained by our algorithm. Finally, we analyze geometric properties of the typical Cox–Voronoi cell and compare them to properties of the typical cell of other well-known classes of tessellations, especially Poisson–Voronoi tessellations. Our results can be applied to stochastic–geometric modelling of networks in telecommunication and life sciences, for example. The lines can then represent roads in urban road systems, blood arteries or filament structures in biological tissues or cells, while the points can be locations of telecommunication equipment or vesicles, respectively.

Generalized vector quasi-equilibrium problems

Abstract: This paper deals with generalized vector quasi-equilibrium problems. Using a so-called nonlinear scalarization function and a fixed point theorem, existence theorems for two classes of generalized vector quasi-equilibrium problems are established.

Cost benefit analysis of series systems with warm standby components and general repair time

Abstract: We study the availability analysis of three different series system configurations with warm standby components and general repair times. The time-to-failure for each of the primary and warm standby components is assumed to be exponentially distributed with respective parameter λ and α. This paper presents a recursive method, using the supplementary variable technique and treating the supplementary variable as the remaining repair time, to develop the steady-state probability distribution of the number of working components in the system. We develop the explicit expressions for the steady-state availability, Av, for three configurations and perform comparisons. For all three configurations, comparisons are made for specific values of distribution parameters and of the cost of the components. The configurations are ranked based on Av and cost/benefit, for three various repair time distributions: exponential, 3-stage Erlang, and deterministic, where benefit is Av.

Cost optimal periodic train scheduling

Abstract: For real world railroad networks, we consider minimizing operational cost of train schedules which depend on choosing different train types of diverse speed and cost. We develop a mixed integer linear programming model for this train scheduling problem. For practical problem sizes, it seems to be impossible to directly solve the model within a reasonable amount of time. However, suitable decomposition leads to much better performance. In the first part of the decomposition, only the train type related constraints stay active. In the second part, using an optimal solution of this relaxation, we select and fix train types and try to generate a train schedule satisfying the remaining constraints. This decomposition idea provides the cornerstone for an algorithm integrating cutting planes and branch-and-bound. We present computational results for railroad networks from Germany and the Netherlands.

On two-stage convex chance constrained problems

Abstract: In this paper we develop approximation algorithms for two-stage convex chance constrained problems. Nemirovski and Shapiro [18] formulated this class of problems and proposed an ellipsoid-like iterative algorithm for the special case where the impact function f(x,h) is bi-affine. We show that this algorithm extends to bi-convex f(x,h) in a fairly straightforward fashion. The complexity of the solution algorithm as well as the quality of its output are functions of the radius r of the largest Euclidean ball that can be inscribed in the polytope defined by a random set of linear inequalities generated by the algorithm [18]. Since the polytope determining r is random, computing r is difficult. Yet, the solution algorithm requiresr as an input. In this paper we provide some guidance for selecting r. We show that the largest value of r is determined by the degree of robust feasibility of the two-stage chance constrained problem – the more robust the problem, the higher one can set the parameter r. Next, we formulate ambiguous two-stage chance constrained problems. In this formulation, the random variables defining the chance constraint are known to have a fixed distribution; however, the decision maker is only able to estimate this distribution to within some error. We construct an algorithm that solves the ambiguous two-stage chance constrained problem when the impact function f(x,h) is bi-affine and the extreme points of a certain “dual” polytope are known explicitly.

Bias and overtaking equilibria for zero-sum continuous-time Markov games

Abstract: This paper deals with continuous-time zero-sum two-person Markov games with denumerable state space, general (Borel) action spaces and possibly unbounded transition and reward/cost rates. We analyze the bias optimality and the weakly overtaking optimality criteria. An example shows that, in contrast to control (or one-player) problems, these criteria are not equivalent for games.

Zero-sum constrained stochastic games with independent state processes

Abstract: We consider a zero-sum stochastic game with side constraints for both players with a special structure. There are two independent controlled Markov chains, one for each player. The transition probabilities of the chain associated with a player as well as the related side constraints depend only on the actions of the corresponding player; the side constraints also depend on the player’s controlled chain. The global cost that player 1 wishes to minimize and that player 2 wishes to maximize, depend however on the actions and Markov chains of both players. We obtain a linear programming formulations that allows to compute the value and saddle point policies for this problem. We illustrate the theoretical results through a zero-sum stochastic game in wireless networks in which each player has power constraints

Staffing decisions for heterogeneous workers with turnover

Abstract: In this paper we consider a firm that employs heterogeneous workers to meet demand for its product or service. Workers differ in their skills, speed, and/or quality, and they randomly leave, or turn over. Each period the firm must decide how many workers of each type to hire or fire in order to meet randomly changing demand forecasts at minimal expense. When the number of workers of each type can by continuously varied, the operational cost is jointly convex in the number of workers of each type, hiring and firing costs are linear, and a random fraction of workers of each type leave in each period, the optimal policy has a simple hire- up-to/fire-down-to structure. However, under the more realistic assumption that the number of workers of each type is discrete, the optimal policy is much more difficult to characterize, and depends on the particular notion of discrete convexity used for the cost function. We explore several different notions of discrete convexity and their impact on structural results for the optimal policy.

The G/M/1 queue revisited

Abstract: The G/M/1 queue is one of the classical models of queueing theory. The goal of this paper is two-fold: (a) To introduce new derivations of some well-known results, and (b) to present some new results for the G/M/1 queue and its variants. In particular, we pay attention to the G/M/1 queue with a set-up time at the start of each busy period, and the G/M/1 queue with exceptional first service.

ɛ-Subdifferentials of Set-valued Maps and ɛ-Weak Pareto Optimality for Multiobjective Optimization

Abstract: In this paper we consider vector optimization problems where objective and constraints are set-valued maps. Optimality conditions in terms of Lagrange-multipliers for an ɛ-weak Pareto minimal point are established in the general case and in the case with nearly subconvexlike data. A comparison with existing results is also given. Our method used a special scalarization function, introduced in optimization by Hiriart-Urruty. Necessary and sufficient conditions for the existence of an ɛ-weak Pareto minimal point are obtained. The relation between the set of all ɛ-weak Pareto minimal points and the set of all weak Pareto minimal points is established. The ɛ-subdifferential formula of the sum of two convex functions is also extended to set-valued maps via well known results of scalar optimization. This result is applied to obtain the Karush–Kuhn–Tucker necessary conditions, for ɛ-weak Pareto minimal points

Locating a maximally complementary solution of the monotone NCP by using non-interior-point smoothing algorithms

Abstract: In this paper we propose a non-interior-point smoothing algorithm for solving the monotone nonlinear complementarity problem (NCP). The proposed algorithm is simpler than many existing non-interior-point smoothing algorithms in the sense that it only needs to solve one system of linear equations and to perform one line search at each iteration. We show that the proposed algorithm is globally convergent under the assumption that the NCP concerned has a nonempty solution set. Such assumption is weaker than those required by most other non-interior-point smoothing algorithms. In particular, we prove that the solution obtained by the proposed algorithm is a maximally complementary solution of the NCP concerned. Preliminary numerical results are reported.

Managing the reputation of an award to motivate performance

Abstract: Managers wish to motivate workers to exert effort. There is large literature on the use of wages and monetary incentives for this purpose, but in practice the “honor” or “prestige” of an award can be a significant motivator as well, unless the award is given so often that its prestige is diluted. The model here focuses on management of the reputation of an award that may or may not have a fixed monetary component. The model is an optimal dynamic control model, so its solution suggests how to manage the award over time. The analysis is interesting because of a “false” steady state that is adjacent to but outside the admissible region and which otherwise has the qualitative properties of a steady state; there are (infinitely many) trajectories converging to it and (infinitely many) trajectories starting arbitrarily close to it. For all initial conditions there are infinitely many candidates for the optimal solution that cannot be evaluated in the standard way. We resolve that problem by proving a new proposition concerning the value of the utility functional when the limit of the Hamiltonian is non-zero. Managerial implications are derived.

Stochastic Control Problems with Delay

Abstract: We consider optimal control problems for systems described by stochastic differential equations with delay. We state conditions for certain classes of such systems under which the stochastic control problems become finite-dimensional. These conditions are illustrated with three applications. First, we solve some linear quadratic problems with delay. Then we find the optimal consumption rate in a financial market with delay. Finally, we solve explicitly a deterministic fluid problem with delay which arises from admission control in ATM communication networks.

Deviation Measures in Linear Two-Stage Stochastic Programming

Abstract: We consider a linear two-stage stochastic program. Whereas optimization in the traditional setting is based solely on expectation, we include risk measures reflecting dispersions of the random objective. Presenting the mean-risk models, we aim to extend existing results for the expectation-based model. In particular, we discuss structural properties such as continuity, differentiability and convexity and address stability issues. Furthermore, we propose algorithmic treatment with a slight variation of the L-shaped method

Control of ruin probabilities by discrete-time investments

Abstract: The control problem of controlling ruin probabilities by investments in a financial market is studied. The insurance business is described by the usual Cramer-Lundberg-type model and the risk driver of the financial market is a compound Poisson process. Conditions for investments to be profitable are derived by means of discrete-time dynamic programming. Moreover Lundberg bounds are established for the controlled model.

The Laurent series, sensitive discount and Blackwell optimality for continuous-time controlled Markov chains

Abstract: This paper gives conditions for the convergence of the Laurent series expansion for a class of continuous-time controlled Markov chains with possibly unbounded reward (or cost) rates and unbounded transition rates That series is then used to study several optimization criteria, including n-discount optimality (for n=−1,0,1,), Blackwell optimality, and the maximization of a certain vector criterion that in particular gives gain and bias optimality

Nonlinear-Programming Reformulation of the Order-Value Optimization problem

Abstract: Order-value optimization (OVO) is a generalization of the minimax problem motivated by decision-making problems under uncertainty and by robust estimation. New optimality conditions for this nonsmooth optimization problem are derived. An equivalent mathematical programming problem with equilibrium constraints is deduced. The relation between OVO and this nonlinear-programming reformulation is studied. Particular attention is given to the relation between local minimizers and stationary points of both problems.

A preference change and discretionary stopping in a consumption and porfolio selection problem

Abstract: We study an optimal consumption-portfolio selection problem in which an economic agent is able to choose a discretionary stopping time in a continuous-time framework. We focus on studying the problem for the case where the agent’s preference changes around the stopping time. We obtain the optimal policy in an explicit form by solving free boundary value problems. If the agent’s coefficient of relative risk aversion becomes higher (lower) after the stopping time, then the optimal policy is to stop as soon as the wealth level touches down (up) to the critical wealth level.

Increasing quasiconcave co-radiant functions with applications in mathematical economics

Abstract: We study increasing quasiconcave functions which are co-radiant. Such functions have frequently been employed in microeconomic analysis. The study is carried out in the contemporary framework of abstract convexity and abstract concavity. Various properties of these functions are derived. In particular we identify a small “natural” infimal generator of the set of all coradiant quasiconcave increasing functions. We use this generator to examine two duality schemes for these functions: classical duality often used in microeconomic analysis and a more recent duality concept. Some possible applications to the theory of production functions and utility functions are discussed.

Hahn-Banach theorems and subgradients of set-valued maps

Abstract: Some new results which generalize the Hahn-Banach theorem from scalar or vector-valued case to set-valued case are obtained. The existence of the Borwein-strong subgradient and Yang-weak subgradient for set-valued maps are also proven. We present a new Lagrange multiplier theorem and a new Sandwich theorem for set-valued maps.

A unified approach to portfolio optimization with linear transaction costs

Abstract: In this paper we study the continuous time optimal portfolio selection problem for an investor with a finite horizon who maximizes expected utility of terminal wealth and faces transaction costs in the capital market. It is well known that, depending on a particular structure of transaction costs, such a problem is formulated and solved within either stochastic singular control or stochastic impulse control framework. In this paper we propose a unified framework, which generalizes the contemporary approaches and is capable to deal with any problem where transaction costs are a linear/piecewise-linear function of the volume of trade. We also discuss some methods for solving numerically the problem within our unified framework.

On multiple simple recourse models

Abstract: We consider multiple simple recourse (MSR) models, both continuous and integer versions, which generalize the corresponding simple recourse (SR) models by allowing for a refined penalty cost structure for individual shortages and surpluses. It will be shown that (convex approximations of) such MSR models can be represented as explicitly specified continuous SR models, and thus can be solved efficiently by existing algorithms.

Portfolio optimization under transaction costs in the CRR model

Abstract: In the CRR model we introduce a transaction cost structure which covers piecewise proportional, fixed and constant costs. For a general utility function we formulate the problem of maximizing the expected utility of terminal wealth as a Markov control problem. An existence result is given and optimal strategies can be described by solutions of the dynamic programming equation. For logarithmic utility we provide detailed solutions in the one-period case and provide examples for the multi-dimensional case and for complex cost structures. For a combination of fixed and proportional costs a fast multi-period algorithm is introduced.

Drop out monotonic rules for sequencing situations

Abstract: This note introduces a new monotonicity property for sequencing situations. A sequencing rule is called drop out monotonic if no player will be worse off whenever one of the players decides to drop out of the queue before processing starts. This intuitively appealing property turns out to be very strong: we show that there is at most one rule satisfying both stability and drop out monotonicity. For the standard model of linear cost functions, the existence of this rule is established.

On essential information in sequential decision processes

Abstract: This paper provides sufficient conditions when certain information about the past of a stochastic decision processes can be ignored by a controller. We illustrate the results with particular applications to queueing control, control of semi-Markov decision processes with iid sojourn times, and uniformization of continuous-time Markov decision processes.

Semi-Markov control processes with unknown holding times distribution under a discounted criterion

Abstract: The paper deals with a class of semi-Markov control models with Borel state and control spaces, possibly unbounded costs, and unknown holding times distribution H. Assuming that H does not depend on state-action pairs, we combine suitable methods of statistical estimation of H with control procedures to construct an asymptotically discounted optimal policy and an optimal stationary policy { f∞}, where f n converges to f∞ in the sense of Schal [12].

Generalized Pinwheel Problem

Abstract: This paper studies a non-preemptive infinite-horizon scheduling problem with a single server and a fixed set of recurring jobs. Each job is characterized by two given positive numbers: job duration and maximum allowable time between the job completion and its next start. We show that for a feasible problem there exists a periodic schedule. We also provide necessary conditions for the feasibility, formulate an algorithm based on dynamic programming, and, since this problem is NP-hard, formulate and study heuristic algorithms. In particular, by applying the theory of Markov Decision Process, we establish natural necessary conditions for feasibility and develop heuristics, called frequency based algorithms, that outperform standard scheduling heuristics.