2004).These models capture, in full generality, the trade-offs among capacity, uncertainty, and customer choice that define RM.Yet, as already noted by Talluri and van Ryzin (1998), the resulting problems are often simply too large to solve exactly.The enduring challenge, therefore, lies not in formulating the "perfect" model but in extracting useful, implementable decision policies from systems that defy exact optimization.This Special Issue on Revenue Management for Complex Systems brings together five contributions that embody precisely this philosophy.Each paper begins with a theoretically sound model that captures the richness of real-world complexity (multidimensional heterogeneity, dynamic decisions under uncertainty, coupled subsystems, and data-driven learning) but then forges a computational path that makes the model tractable and actionable.Although they operate across diverse domains from pricing consumer products to managing transportation, logistics, and mobility networks, they share a unifying spirit: to turn complexity into structure, and structure into insight. Dynamic Nonlinear Pricing under Multiunit DemandWhen customers differ not only in their willingness to pay but also in the quantity they demand, even the simplest dynamic pricing problems become analytically

Abstract The paper deals with a modern approach of stochastic dominance in portfolio optimization. Since the distribution of returns is often just estimated from data, we look for the worst-case distribution that differs from the empirical distribution by no more than some prescribed value. First, we define in what sense the distribution is the worst one for stochastic dominance. Then, using Wasserstein distance, we derive a reformulation for robust second-order stochastic dominance and find the worst-case distribution as the optimal solution of a non-linear optimization problem. Finally, we derive programs to maximize an objective function over the weights of the portfolio with the robust stochastic dominance condition in constraints. We consider robustness in returns for second-order stochastic dominance. We apply all derived optimization programs to real-life data, specifically to returns of assets captured by the Dow Jones Industrial Average, and analyze the problems in detail using optimal solutions of optimization programs with multiple setups. The empirical analysis proceeded with an out-of-sample evaluation of portfolios formulated through the robust optimization program, employing a moving window methodology. The findings of this study indicate that for some of the values of $$\varepsilon $$ ε the robustified portfolios consistently out-of-sample outperform those derived from the non-robust optimization approach.

Abstract We consider the problem of collecting and processing waste material. At a production facility, a known amount of inventory is required for production (e.g., paper) for every period. Instead of new material, the facility relies on collected and processed waste material (e.g., paper waste). This material is collected from regional waste collection locations. The amount of waste material per location is uncertain, as is the quality of the collected waste, i.e., the resulting inventory when processing the material. If the inventory is insufficient at the end of a period, costly new material must be bought. Each period, decisions are made about how much waste material to collect from which location and how to route the collection vehicles accordingly. Ideally, inventory is built to hedge against quality uncertainty and to ensure efficient routing operations in future periods. We propose a stochastic lookahead method that samples a set of scenarios and solves a simplified two-stage stochastic program in every period. We show the value of our method for two case studies, one based on real-world data from Sachsen-Anhalt, Germany, and one from the literature with data from the United Kingdom. We further conduct a detailed analysis of our method and the problem characteristics. The results show that our method effectively anticipates all sources of uncertainty, reducing cost significantly compared to benchmark policies. This superior performance is due to appropriate state-dependent supplier selection that considers the percentage of material loss, available material, and routing cost for current and future periods.

Abstract Airports and airlines share a common interest in providing passengers with fast and convenient access to airports. With the introduction of air taxis as a new means of transport, the opportunity will soon emerge to introduce air taxi shuttle services to airports. In this context, for each individual airport the question arises where the facilities for landing, ground handling and take-off, also called vertiports, should be located in the area surrounding the airport. We formulate the problem as a hub location problem with multiple allocation, with the objective to maximize the number of passengers who choose to reach the airport by air taxi. In our model, we explicitly incorporate passengers’ choice behavior with respect to available means of transport using a multinomial logit model. We linearize the problem and provide guidance on the steps needed to use the model in practice. In an extensive case study, we apply the model to real data for Munich Airport, perform a stated preference study to parameterize the multinomial logit model, suggest locations for vertiports throughout Bavaria as a result, and investigate to what extent the introduction of air taxi shuttle flights can reduce travel times to Munich Airport.

Multiobjective linear fractional programming is useful to model multiobjective problems where all or some of the objective functions are a ratio or proportion of one linear/affine function to another linear/affine function. In practice, many of such problems include integer variables. If the weighted-sum scalarization is used to compute efficient solutions to the multiobjective problem, then the scalar problem to be solved for each weight vector turns out to be a weighted sum-of-ratios. There are several algorithms reported in the literature to optimize weighted sum-of-ratios, but almost all of them cannot deal with integer variables. In this paper we propose a Branch & Cut algorithm to optimize weighted-sums of the objective functions in multiobjective mixed integer fractional programming (MOMIFP). Several theoretical properties that support the algorithm are presented and proved. Computational experiments with randomly generated general problems are presented and discussed, which show that the algorithm is able to deal with practical MOMIFP problems.