Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

WB52

4 - Determining Catastrophe Insurance Subsidies under Social Network Influence Shenming Song, Tsinghua University, Beijing, China, Chen (Mavis) Wang Catastrophe insurance is regarded as an effective financial mechanism that transfers post-disaster losses to proactive preparedness investments, despite many practical challenges. We consider a sequential game with three types of players, the government, a private insurer, and a community of networked policy holders, with which we examine the tradeoffs between government subsidies for insurance premiums and post-event subsidies under different characteristics of social network influence. We also explore the benefits of a multi-year insurance contract over consecutive single-year contracts.

programming problem, and use a Lagrangian Relaxation approach to derive heuristic allocation policies. We evaluate the flexibility and resilience that emerge from these policies, and analyze how various environmental factors impact performance. 6 - Dynamic Project Expediting in Stochastic Networks Luca Bertazzi, University of Brescia, Brescia, 25122, Italy, Riccardo Mogre A project manager is in charge of a project. At each time, she needs to decide the effort level to invest in the project. The progress made on the project is random, constituting disruptions or efficiency problems. We formulate a stochastic discrete dynamic programming model for this problem and design an exact algorithm to find an optimal policy. Computational results show that this algorithm is more efficient than the classical exact algorithms Value iteration, Policy iteration and Linear programming. n WB53 North Bldg 232A Joint Session AMD/Practice Curated: Auctions and Mechanism Design Applications Sponsored: Auction and Marketing Design Sponsored Session Chair: Michael O. Ball, University of Maryland-College Park, College Park, MD, 20742-1815, United States 1 - Quantity-contingent Auctions and Allocation of Airport Slots Yulin Liu, University of California-Berkeley, 107D McLaughlin Hall, Berkeley, CA, 94720, United States, Michael O. Ball, Alexander Estes, Mark M. Hansen We define and investigate quantity-contingent auctions, which can be used when there exist multiple units of a single product and the value of a set of units depends on the total quantity sold. A quantity-contingent auction determines both the number of items sold and an allocation of items to bidders. We focus on auctions that allocate airport arrival and departure slots. We propose a continuous model and an integer programming model for the associated winner determination problem. Using these models, we perform computational experiments that lend insights into the properties of the quantity-contingent auction. 2 - Strategic Timing and Pricing in On-demand Platforms Mustafa Dogan, Carnegie Mellon University, 5000 Forbes Avenue, Pittsburgh, PA, 15213, United States, Alexandre Jacquillat, Vibhanshu Abhishek We design a dynamic pricing and allocation mechanism for service provision in an on-demand platform facing demand stochasticity, heterogeneity across price- sensitive and time-sensitive customers, and information asymmetry. Time is a strategic device to dynamically manage the demand-capacity imbalances; and to provide discriminated service levels. Results suggest that the optimal mechanism depends on the strength of customer heterogeneity and the time preferences of price-sensitive customers. The proposed mechanism increases platform profits as compared to surge pricing policies, and can even provide a Pareto improvement. We also show that higher demand may trigger lower price. 3 - How Efficient is CTOP? Alexander Estes, University of Maryland, Beltsville, MD, 20705, United States, Farzad Daneshgar, Michael O. Ball The Federal Aviation Administration sometimes issues traffic management initiatives that restrict the amount of traffic allowed to enter some region of airspace. The Collaborative Trajectory Options Program allows flight operators to specify alternative routes for their flights. The FAA uses the options to more efficiently implement the traffic management initiative. In this work, we compare the efficiency of the current CTOP allocation process and that of a theoretical system-optimal allocation process. 4 - Randomized Mechanism to Coordinate Carriers in Retail Logistics Paul Karaenke, Technical University of Munich, Department of Informatics (I18), Boltzmannstr 3, Munich, 85748, Germany, Martin Bichler, Stefan Minner We consider the problem of slot booking by independent carriers at several warehouses and investigate recent developments in the design of electronic market mechanisms promising to address both computational and strategic complexity. Relax-and-round mechanisms describe a class of approximation mechanisms that is truthful in expectation and runs in polynomial time. While the solution quality of these mechanisms is low, we introduce a variant able to solve real-world problem sizes with high solution quality while still being incentive-compatible. We compare these mechanisms to core-selecting auctions, which are not incentive-compatible, but provide stable outcomes with respect to the bids.

n WB52 North Bldg 231C Practice - Nonlinear and Dynamic Programming

for Applications Contributed Session

Chair: Luca Bertazzi, University of Brescia, Brescia, 25122, Italy 1 - A One-phase Interior Point Method for Nonconvex Optimization Oliver Hinder, Stanford, Stanford, CA, 94305, United States, Yinyu Ye The work of Wachter and Biegler suggests that infeasible-start interior point methods (IPMs) developed for linear programming cannot be adapted to nonlinear optimization without significant modification, i.e., using a two-phase or penalty method. We propose an IPM that, by careful initialization and updates of the slack variables, is guaranteed to find a first-order certificate of local infeasibility, local optimality, or unboundedness of the (shifted) feasible region. Our algorithm with closely resembles successful algorithms from linear programming. Experiments also indicate superior robustness and infeasibility detection compared with IPOPT. 2 - Power System Optimization through Joint Placement of Phasor and Flow Measurements Vahidhossein Khiabani, Assistant Professor, Middle Tennessee State University, Murfreesboro, TN, 37132, United States We proposed a mathematical method for joint optimal placement of phasor and conventional flow measurements considering the conflicting objectives of Phasor Measurement Units (PMU’s) deployment cost, and system reliability. The model places PMUs on a network with a pre-existing conventional measurements, considering zero injection measurements with the goal of complete observability of the system. The proposed approach is tested on IEEE standard bus systems and compared with some models in literature. 3 - Algorithm for Evolutionarily Stable Strategies against Pure Mutations Sam Ganzfried, Ganzfried Research, Miami Beach, FL, 33139, United States Evolutionarily stable strategy (ESS) is an important solution concept in game theory which has been applied frequently to biology and even cancer. Finding such a strategy has been shown to be difficult from a theoretical complexity perspective. Informally an ESS is a strategy that if followed by the population cannot be taken over by a mutation strategy. We present an algorithm for the case where mutations are restricted to pure strategies. This is the first positive result for computation of ESS, as all prior results are computational hardness and no prior algorithms have been presented. 4 - Optimization Tools for the Management of Electric Vehicles in Electrical Networks Rafael Zarate-Minano, Associate Professor, University of Castilla- La Mancha, EIMIA, Plaza Manuel Meca 1, Almaden, 13400, Spain, Alberto Flores, Miguel Carrion, Ruth Dominguez The generalized use of the electric vehicle in industralized countries could become a reality during the next decade. In this context, it is necessary to carefully anaylize different aspects of the interaction between these vehicles and electrical networks. This work explores the combined application of optimization methods and homotopy tecniques for the modeling of such interaction. The ability of this combination of tecniques to study aspects such as the adecuacy and the operation of the existing transmission and distribution networks in a context of large-scale integration of the electric vehicle is discussed. 5 - Dynamic Task Allocation with Learning and Forgetting Thomas Vossen, University of Colorado-Boulder, Leeds School of Business, UCB419, Boulder, CO, 80309, United States, Peter Letmathe We consider a setting where tasks arrive randomly over time for possible processing. Incoming tasks can be allocated to (human) resources, whose productivity depends on the number of tasks processed by the resource before (learning) and is impacted by changes in the workforce over time (forgetting). We formulate the task allocation problem as a weakly coupled stochastic dynamic

460