Decomposition for adjustable robust linear optimization subject to uncertainty polytope

We present in this paper a general decomposition framework to solve exactly adjustable robust linear optimization problems subject to polytope uncertainty. Our approach is based on replacing the polytope by the set of its extreme points and generating the extreme points on the fly within row generation or column-and-row generation algorithms. The novelty of our approach lies in formulating the separation problem as a feasibility problem instead of a max–min problem as done in recent works. Applying the Farkas lemma, we can reformulate the separation problem as a bilinear program, which is then linearized to obtained a mixed-integer linear programming formulation. We compare the two algorithms on a robust telecommunications network design under demand uncertainty and budgeted uncertainty polytope. Our results show that the relative performance of the algorithms depend on whether the budget is integer or fractional.     

A note on issues of over-conservatism in robust optimization with cost uncertainty

We identify and discuss issues of hidden over-conservatism in robust linear optimization, when the uncertainty set is polyhedral with a budget of uncertainty constraint. The decision-maker selects the budget of uncertainty to reflect his degree of risk aversion, i.e. the maximum number of uncertain parameters that can take their worst-case value. In the first setting, the cost coefficients of the linear programming problem are uncertain, as is the case in portfolio management with random stock returns. We provide an example where, for moderate values of the budget, the optimal solution becomes independent of the nominal values of the parameters, i.e. is completely disconnected from its nominal counterpart, and discuss why this happens. The second setting focusses on linear optimization with uncertain upper bounds on the decision variables, which has applications in revenue management with uncertain demand and can be rewritten as a piecewise linear problem with cost uncertainty. We show in an example that it is possible to have more demand parameters equal their worst-case value than what is allowed by the budget of uncertainty, although the robust formulation is correct. We explain this apparent paradox.     

Robust shift generation in workforce planning

In this paper we apply robust optimization techniques to the shift generation problem in workforce planning. At the time that the shifts are generated, there is often much uncertainty in the workload predictions. We propose a model to generate shifts that are robust against this uncertainty. An adversarial approach is used to solve the resulting robust optimization model. In each iteration an integer nonlinear knapsack problem is solved to calculate the worst case workload scenario. We apply the approach to generate shifts in a real-life Air Traffic Controller workforce planning problem. The numerical results show the value of our approach.     

An implicit enumeration algorithm for the passenger service planning problem: Application to the Taiwan Railways Administration line

In a passenger railroad system, the service planning problem determines the train stopping strategy, taking into consideration multiple train classes and customer origin–destination (OD) demand, to maximize the short-term operational profit of a rail company or the satisfaction levels of the passengers. The service plan is traditionally decided by rule of thumb, an approach that leaves much room for improvement. To systematically analyze this problem, we propose an integer program approach to determine the optimal service plan for a rail company. The formulated problem has a complex solution space, and commonly used commercial optimization packages are currently incapable of solving this problem efficiently, especially when problems of realistic sizes are considered. Therefore, we develop an implicit enumeration algorithm that incorporates intelligent branching and effective bounding strategies so that the solution space of this integer program can be explored efficiently. The numerical results show that the proposed implicit enumeration algorithm can solve real-world problems and can obtain service plans that are at least as good as those developed by the rail company.     

Adaptive and robust radiation therapy optimization for lung cancer

A previous approach to robust intensity-modulated radiation therapy (IMRT) treatment planning for moving tumors in the lung involves solving a single planning problem before the start of treatment and using the resulting solution in all of the subsequent treatment sessions. In this paper, we develop an adaptive robust optimization approach to IMRT treatment planning for lung cancer, where information gathered in prior treatment sessions is used to update the uncertainty set and guide the reoptimization of the treatment for the next session. Such an approach allows for the estimate of the uncertain effect to improve as the treatment goes on and represents a generalization of existing robust optimization and adaptive radiation therapy methodologies. Our method is computationally tractable, as it involves solving a sequence of linear optimization problems. We present computational results for a lung cancer patient case and show that using our adaptive robust method, it is possible to attain an improvement over the traditional robust approach in both tumor coverage and organ sparing simultaneously. We also prove that under certain conditions our adaptive robust method is asymptotically optimal, which provides insight into the performance observed in our computational study. The essence of our method – solving a sequence of single-stage robust optimization problems, with the uncertainty set updated each time – can potentially be applied to other problems that involve multi-stage decisions to be made under uncertainty.     

An exact algorithm for the Capacitated Total Quantity Discount Problem

In this paper we analyze the procurement problem of a company that needs to purchase a number of products from a set of suppliers to satisfy demand. The suppliers offer total quantity discounts and the company aims at selecting a set of suppliers so to satisfy product demand at minimum purchasing cost. The problem, known as Total Quantity Discount Problem (TQDP), is strongly NP-hard. We study different families of valid inequalities and provide a branch-and-cut approach to solve the capacitated variant of the problem (Capacitated TQDP) where the quantity available for a product from a supplier is limited. A hybrid algorithm, called HELP (Heuristic Enhancement from LP), is used to provide an initial feasible solution to the exact approach. HELP exploits information provided by the continuous relaxation problem to construct neighborhoods optimally searched through the solution of mixed integer subproblems. A streamlined version of the proposed exact method can optimally solve in a reasonable amount of time instances with up to 100 suppliers and 500 products, and largely outperforms an existing approach available in the literature and CPLEX 12.2 that frequently runs out of memory before completing the search.     

Robust location of new housing developments using a choice model

We consider the issue of choosing a subset of locations to construct new housing developments maximizing the satisfaction of potential buyers, which has not been previously studied in the literature. The allocation of demands to the selected locations is modeled by a choice model, based on the distance to the location, real-estate prices and incomes. We study two robust counterparts of the optimal location problem, where uncertainty lies on demand volumes for the first one, and on customer preferences for the second one. In both cases, the parameters subject to uncertainty appear both in the objective function and constraints. The second robust model combines a scenario-based approach with nominal, price-centric and distance-centric scenarios on customers preferences, and an uncertainty budget approach that limits the number of customers that can deviate from the nominal scenario. We show that the subproblem of finding the worst-case deviation of parameters subject to uncertainty is tractable and leads to linear formulations of the robust problem. Computational experiments conducted on instances of the Paris region show that the average loss of value of the robust solution is reasonably low when compared to the optimal solution of deviated instances. We also derive insights for the new housing development issue.     

Short sales in Log-robust portfolio management

This paper extends the Log-robust portfolio management approach to the case with short sales, i.e., the case where the manager can sell shares he does not yet own. We model the continuously compounded rates of return, which have been established in the literature as the true drivers of uncertainty, as uncertain parameters belonging to polyhedral uncertainty sets, and maximize the worst-case portfolio wealth over that set in a one-period setting. The degree of the manager’s aversion to ambiguity is incorporated through a single, intuitive parameter, which determines the size of the uncertainty set. The presence of short-selling requires the development of problem-specific techniques, because the optimization problem is not convex. In the case where assets are independent, we show that the robust optimization problem can be solved exactly as a series of linear programming problems; as a result, the approach remains tractable for large numbers of assets. We also provide insights into the structure of the optimal solution. In the case of correlated assets, we develop and test a heuristic where correlation is maintained only between assets invested in. In computational experiments, the proposed approach exhibits superior performance to that of the traditional robust approach.     

Robust combinatorial optimization with variable budgeted uncertainty

We introduce a new model for robust combinatorial optimization where the uncertain parameters belong to the image of multifunctions of the problem variables. In particular, we study the variable budgeted uncertainty, an extension of the budgeted uncertainty introduced by Bertsimas and Sim. Variable budgeted uncertainty can provide the same probabilistic guarantee as the budgeted uncertainty while being less conservative for vectors with few non-zero components. The feasibility set of the resulting optimization problem is in general non-convex so that we propose a mixed-integer programming reformulation for the problem, based on the dualization technique often used in robust linear programming. We show how to extend these results to non-binary variables and to more general multifunctions involving uncertainty set defined by conic constraints that are affine in the problem variables. We present a computational comparison of the budgeted uncertainty and the variable budgeted uncertainty on the robust knapsack problem. The experiments show a reduction of the price of robustness by an average factor of 18 %.     

Robust ranking and portfolio optimization

The portfolio optimization problem has attracted researchers from many disciplines to resolve the issue of poor out-of-sample performance due to estimation errors in the expected returns. A practical method for portfolio construction is to use assets’ ordering information, expressed in the form of preferences over the stocks, instead of the exact expected returns. Due to the fact that the ranking itself is often described with uncertainty, we introduce a generic robust ranking model and apply it to portfolio optimization. In this problem, there are n objects whose ranking is in a discrete uncertainty set. We want to find a weight vector that maximizes some generic objective function for the worst realization of the ranking. This robust ranking problem is a mixed integer minimax problem and is very difficult to solve in general. To solve this robust ranking problem, we apply the constraint generation method, where constraints are efficiently generated by solving a network flow problem. For empirical tests, we use post-earnings-announcement drifts to obtain ranking uncertainty sets for the stocks in the DJIA index. We demonstrate that our robust portfolios produce smaller risk compared to their non-robust counterparts.     

Multi-source facility location–allocation and inventory problem

We consider a joint facility location–allocation and inventory problem that incorporates multiple sources of warehouses. The problem is motivated by a real situation faced by a multinational applied chemistry company. In this problem, multiple products are produced in several plants. Warehouse can be replenished by several plants together because of capabilities and capacities of plants. Each customer in this problem has stochastic demand and certain amount of safety stock must be maintained in warehouses so as to achieve certain customer service level. The problem is to determine number and locations of warehouses, allocation of customers demand and inventory levels of warehouses. The objective is to minimize the expected total cost with the satisfaction of desired demand weighted average customer lead time and desired cycle service level. The problem is formulated as a mixed integer nonlinear programming model. Utilizing approximation and transformation techniques, we develop an iterative heuristic method for the problem. An experiment study shows that the proposed procedure performs well in comparison with a lower bound.     

Supply chain network downsizing with product line pruning using a new demand substitution

This paper presents an optimization model for downsizing a multi-product supply chain facing bankruptcy risk, where multi-functional production facilities are shared for producing a group of substitutable products. In order to determine the potential demand after discontinuation of certain product types, the proposed mixed integer programming model incorporates a new general formulation of demand substitution, which allows arbitrary demand diversion and arbitrary replacement rates between products under investigation. The new substitution formulation enables considering uneven substitutions for supply chain network design and therefore it extends currently used substitution approaches. A series of systematically generated downsizing cases confirm the validity of the here proposed approach, and illustrate how downsized chains can successfully adopt to the new circumstances.     

A continuous DC programming approach to the strategic supply chain design problem from qualified partner set

We present a new continuous approach based on the DC (difference of convex functions) programming and DC algorithms (DCA) to the problem of supply chain design at the strategic level when production of a new market opportunity has to be launched among a set of qualified partners. A well known formulation of this problem is the mixed integer linear program. In this paper, we reformulate this problem as a DC program by using an exact penalty technique. The proposed algorithm is a combination of DCA and Branch and Bound scheme. It works in a continuous domain but provides mixed integer solutions. Numerical simulations on many empirical data sets show the efficiency of our approach with respect to the standard Branch and Bound algorithm.     

A dynamic multi-objective location-allocation model for search and rescue assets

This paper presents a dynamic multi-objective mixed integer linear programming model to optimize the location and allocation of search and rescue (SAR) boats and helicopters to enhance the performance of maritime SAR missions. Our model incorporates simulated incident scenarios to account for demand uncertainty and allows relocation of vessels seasonally. We define three objectives as responding to incidents within a critical time, generating a balanced workload distribution among vessels of various types, and minimizing costs associated with operations and vessel relocations. Implementing a goal programming approach, we solve the problem for various objective function term weights and compare the performance of each solution with respect to 10 different metrics. Using historical incident datasets for the Aegean Sea, we show that the proposed model and solution approach can significantly improve the SAR performance and provide decision support for planners in developing effective and efficient resource location-allocation schemes.     

Dynamic models of production with multiple operations and general processing times

Traditional integer programming model formulations for job-shops and flow-shops do not easily account for characteristics common to high-technology manufacturing such as high-volume semiconductor manufacturing. These characteristics are: (1) products (wafers) are processed by the same machine type more than once during the operation sequence, (2) many lots of similar type are run, and (3) there can be multiple machines of the same type. In this paper, we present two new integer programming formulations which easily account for these characteristics. The approach is based on restricting the allowed domain of events for the start of lot processing. The first model restricts production starts to the beginning of a planning time period. The second model uses a special time grid at each operation with width equal to the processing time, and allows starts to be scheduled at the grid points. In an example problem replicating a high-volume wafer fabrication process, it is shown that it is computationally practical to obtain solutions for the restricted start models where it is not computationally possible for the traditional integer programming model formulations.     

A robust optimization approach to dispatching technicians under stochastic service times

We consider the problem of dispatching technicians to service/repair geographically distributed equipment. This problem can be cast as a vehicle routing problem with time windows, where customers expect fast response and small delays. Estimates of the service time, however, can be subject to a significant amount of uncertainty due to misdiagnosis of the reason for failure or surprises during repair. It is therefore crucial to develop routes for the technicians that would be less sensitive to substantial deviations from estimated service times. In this paper we propose a robust optimization model for the vehicle routing problem with soft time windows and service time uncertainty and solve real-world instances with a branch and price method. We evaluate the efficiency of the approach through computational experiments on real industry routing data.     

一种基于协同信息的团队伙伴选择方法

针对基于协同信息的团队伙伴选择问题,提出了一种决策分析方法。首先,给出了伙伴间的协同关系及基于协同信息的团队伙伴选择问题的描述;然后,构建了基于协同信息的团队伙伴选择的数学模型,该模型属于0-1二次整数规划问题,也是NP-hard问题,为了求解该问题,简要阐述了将0-1二次整数规划问题转化为0-1线性整数规划问题的方法;最后,通过一个实例分析说明了本文提出方法的可行性和有效性。     

Dynamic pricing and inventory control: robust vs. stochastic uncertainty models—a computational study

In this paper, we consider a variety of models for dealing with demand uncertainty for a joint dynamic pricing and inventory control problem in a make-to-stock manufacturing system. We consider a multi-product capacitated, dynamic setting, where demand depends linearly on the price. Our goal is to address demand uncertainty using various robust and stochastic optimization approaches. For each of these approaches, we first introduce closed-loop formulations (adjustable robust and dynamic programming), where decisions for a given time period are made at the beginning of the time period, and uncertainty unfolds as time evolves. We then describe models in an open-loop setting, where decisions for the entire time horizon must be made at time zero. We conclude that the affine adjustable robust approach performs well (when compared to the other approaches such as dynamic programming, stochastic programming and robust open loop approaches) in terms of realized profits and protection against constraint violation while at the same time it is computationally tractable. Furthermore, we compare the complexity of these models and discuss some insights on a numerical example.     

Robust discrete optimization and network flows

We propose an approach to address data uncertainty for discrete optimization and network flow problems that allows controlling the degree of conservatism of the solution, and is computationally tractable both practically and theoretically. In particular, when both the cost coefficients and the data in the constraints of an integer programming problem are subject to uncertainty, we propose a robust integer programming problem of moderately larger size that allows controlling the degree of conservatism of the solution in terms of probabilistic bounds on constraint violation. When only the cost coefficients are subject to uncertainty and the problem is a 0–1 discrete optimization problem on n variables, then we solve the robust counterpart by solving at most n+1 instances of the original problem. Thus, the robust counterpart of a polynomially solvable 0–1 discrete optimization problem remains polynomially solvable. In particular, robust matching, spanning tree, shortest path, matroid intersection, etc. are polynomially solvable. We also show that the robust counterpart of an NP-hard -approximable 0–1 discrete optimization problem, remains -approximable. Finally, we propose an algorithm for robust network flows that solves the robust counterpart by solving a polynomial number of nominal minimum cost flow problems in a modified network. The research of the author was partially supported by the Singapore-MIT alliance.The research of the author is supported by a graduate scholarship from the National University of Singapore.Mathematics Subject Classification (2000): 90C10, 90C15     

Protecting the data-driven newsvendor against rare events: a correction-term approach

We propose an approach to the data-driven newsvendor problem that incorporates a correction factor to account for rare events, when the decision-maker has few historical data points at his disposal but knows the range of the demand. This mitigates a weakness of pure data-driven methodologies, specifically, the fact that they under-protect the system against tail events, which are in general under-observed in the empirical demand distribution. We test the approach in extensive computational experiments and provide a summary table of the numerical experiments to help the decision maker gain further insights.