Knowledge that Transforms

Over the last two decades, robust optimization has emerged as a computationally attractive approach to formulate and solve single-stage decision problems affected by uncertainty. More recently, robust optimization has been successfully applied to multistage problems with continuous recourse. This paper takes a step toward extending the robust optimization methodology to problems with integer recourse, which have largely resisted solution so far. To this end, we approximate two-stage robust binary programs by their corresponding K-adaptability problems, in which the decision maker precommits to K second-stage policies, here -and-now, and implements the best of these policies once the uncertain parameters are observed. We study the approximation quality and the computational complexity of the K-adaptability problem, and we propose two mixed-integer linear programming reformulations that can be solved with off-the-shelf software. We demonstrate the effectiveness of our reformulations for stylized instances of supply chain design, route planning, and capital budgeting problems.

This paper addresses two concerns with the state of the art in network revenue management with dependent demands. The first concern is that the basic attraction model (BAM), of which the multinomial logit (MNL) model is a special case, tends to overestimate demand recapture in practice. The second concern is that the choice-based deterministic linear program, currently in use to derive heuristics for the stochastic network revenue management problem, has an exponential number of variables. We introduce a generalized attraction model (GAM) that allows for partial demand dependencies ranging from the BAM to the independent demand model (IDM). We also provide an axiomatic justification for the GAM and a method to estimate its parameters. As a choice model, the GAM is of practical interest because of its flexibility to adjust product-specific recapture. Our second contribution is a new formulation called the sales-based linear program (SBLP) that works for the GAM. This formulation avoids the exponential number of variables in the earlier choice-based network RM (revenue management) approaches and is essentially the same size as the well-known LP formulation for the IDM. The SBLP should be of interest to revenue managers because it makes choice-based network RM problems tractable to solve. In addition, the SBLP formulation yields new insights into the assortment problem that arises when capacities are infinite. Together these contributions move forward the state of the art for network revenue management under customer choice and competition.

We formalize the reference point adaptation process by relating it to a way people perceive prior gains and losses. We then develop a dynamic trading model with reference point adaptation and loss aversion, and derive its semi-analytical solution. The derived optimal stock holding has an asymmetric V-shaped form with respect to prior outcomes, and the related sensitivities are directly determined by the sensitivities of reference point shifts with respect to the outcomes. We also find that the effects of reference point adaptation can be used to shed light on some well documented trading patterns, e.g., house money, break even, and disposition effects.

The production routing problem (PRP) is a generalization of the inventory routing problem and concerns the production and distribution of a single product from a production plant to multiple customers using capacitated vehicles in a discrete- and finite-time horizon. In this study, we consider the stochastic PRP with demand uncertainty in two-stage and multistage decision processes. The decisions in the first stage include production setups and customer visit schedules, while the production and delivery quantities are determined in the subsequent stages. We introduce formulations for the two problems, which can be solved by a branch-and-cut algorithm. To handle a large number of scenarios, we propose a Benders decomposition approach, which is implemented in a single branch-and-bound tree and enhanced through lower-bound lifting inequalities, scenario group cuts, and Pareto-optimal cuts. For the multistage problem, we also use a warm start procedure that relies on the solution of the simpler two-stage problem. Finally, we exploit the reoptimization capabilities of Benders decomposition in a sample average approximation method for the two-stage problem and in a rollout algorithm for the multistage problem. Computational experiments show that instances of realistic size can be solved to optimality for the two-stage and multistage problems, and that Benders decomposition provides significant speedups compared to a classical branch-and-cut algorithm.

Debris management is one of the most time consuming and complicated activities among post-disaster operations. Debris clearance is aimed at pushing the debris to the sides of the roads so that relief distribution and search-and-rescue operations can be maintained in a timely manner. Given the limited resources, uncertainty, and urgency during disaster response, efficient and effective planning of debris clearance to achieve connectivity between relief demand and supply is important. In this paper, we define the stochastic debris clearance problem (SDCP), which captures post-disaster situations where the limited information on the debris amounts along the roads is updated as clearance activities proceed. The main decision in SDCP is to determine a sequence of roads to clear in each period such that benefit accrued by satisfying relief demand is maximized. To solve SDCP to optimality, we develop a partially observable Markov decision process model. We then propose a heuristic based on a continuous-time approximation, and we further reduce the computational burden by applying a limited look ahead on the search tree and heuristic pruning. The performance of these approaches is tested on randomly generated instances that reflect various geographical and information settings, and instances based on a real-world earthquake scenario. The results of these experiments underline the importance of applying a stochastic approach and indicate significant improvements over heuristics that mimic the current practice for debris clearance.

Consider the newsvendor model, but under the assumption that the underlying demand distribution is not known as part of the input. Instead, the only information available is a random, independent sample drawn from the demand distribution. This paper analyzes the sample average approximation (SAA) approach for the data-driven newsvendor problem. We obtain a new analytical bound on the probability that the relative regret of the SAA solution exceeds a threshold. This bound is significantly tighter than existing bounds, and it matches the empirical accuracy of the SAA solution observed in extensive computational experiments. This bound reveals that the demand distribution’s weighted mean spread affects the accuracy of the SAA heuristic.

We study a dynamic pricing problem with finite inventory and parametric uncertainty on the demand distribution. Products are sold during selling seasons of finite length, and inventory that is unsold at the end of a selling season perishes. The goal of the seller is to determine a pricing strategy that maximizes the expected revenue. Inference on the unknown parameters is made by maximum-likelihood estimation. We show that this problem satisfies an endogenous learning property, which means that the unknown parameters are learned on the fly if the chosen selling prices are sufficiently close to the optimal ones. We show that a small modification to the certainty equivalent pricing strategy—which always chooses the optimal price w.r.t. current parameter estimates—satisfies Regret(T) = O(log2(T)), where Regret(T) measures the expected cumulative revenue loss w.r.t. a clairvoyant who knows the demand distribution. We complement this upper bound by showing an instance for which the regret of any pricing policy satisfies Ω(log T).

We consider a non-stationary variant of a sequential stochastic optimization problem, in which the underlying cost functions may change along the horizon. We propose a measure, termed variation budget, that controls the extent of said change, and study how restrictions on this budget impact achievable performance. We identify sharp conditions under which it is possible to achieve long-run average optimality and more refined performance measures such as rate optimality that fully characterize the complexity of such problems. In doing so, we also establish a strong connection between two rather disparate strands of literature: (1) adversarial online convex optimization and (2) the more traditional stochastic approximation paradigm (couched in a non-stationary setting). This connection is the key to deriving well-performing policies in the latter, by leveraging structure of optimal policies in the former. Finally, tight bounds on the minimax regret allow us to quantify the “price of non-stationarity,” which mathematically captures the added complexity embedded in a temporally changing environment versus a stationary one.