Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

TE35

mechanism should satisfy some socially acceptable properties, for example, unanimity and strategy-proofness. It is well known that an RSCF is unanimous and strategy-proof if and only if it is a random dictatorship. This leads to a roadblock for designing socially acceptable and truthful voting mechanisms. We construct a normalized scoring-based rule that is unanimous and weakly strategy- proof and, therefore, it is not dictatorial. It is fast to compute and easy to implement. n TE34 North Bldg 223 4:35 - 5:20 Responsive Learning Technologies/ 5:20 - 6:05 Lumina Decision Systems Emerging Topic: Technology Tutorials Emerging Topic Session 1 - Online Games to Teach Operations and Supply Chain Management Littlefield Technologies, the Supply Chain Game, and the Sourcing Game are online competitive assignments used to teach topics including process analysis, inventory control, supply chain management, and sourcing and purchasing. We will describe the games’ learning objectives, typical assignments, and actual game results. 2 - How to Engage with Your Clients for More Effective Analytics Max Henrion, Lumina Decision Systems, Inc, Campbell, CA, United States In this workshop, Max Henrion will show how you can use Analytica as a key aid for more effective conversations with your clients, including how to: Draw influence diagrams to help clients articulate their real objectives and decisions, and to work with them to frame and scope problems. Use sensitivity analysis to help your clients understand what data and assumptions really matter and why. Employ agile modeling methods to build decision tools that end users find usable and useful. Design compelling visualizations to provide your clients a basis for informed and confident decisions. Practicing analysts and modelers will find these methods effective for improving client engagement in conjunction with almost any analytics software, but Analytica is unique in offering features designed specifically to support this approach to interactive modeling. Active participants will receive a free 12-month Analytica license. If you already have Analytica, you can give it to a colleague. Samuel C. Wood, Responsive Learning Techologies, 4546 El Camino Real, #243, Los Altos, CA, 94022, United States Chair: Reed Harder, Dartmouth, Hanover, NH, 03755, United States 1 - Machine Learning, Discrete Games, and Predictive Models of Entry in the Airline Industry Kang Hua Cao, Hong Kong Baptist University, WLB 517, 34 Renfrew Road, Kowloon Tong, Hong Kong, Lei Kang, Chia-Mei Liu, Vikrant Vaze In this paper, we intend to widen connections, and develop further relationships, between machine learning and economic theory. Existing forecasting methods use the reduced-form approach and maintain the so-called fixed network assumption. Ignoring the strategic effects in forecasting can lead to biased results. We implement supervised machine learning methods to infer the general form of the best response functions and the conditional choice probabilities of the game. Then, we apply the model and methods developed to the U.S. airline industry, to predict the probabilities involved in an airline company entering or exiting a market. 2 - Game-theoretic Analysis of Reallocation Mechanisms of Airport Landing Slots We study reallocation mechanisms in which airlines can exchange landing slots amongst each other in the context of ground delay programs. Such mechanisms should not only be able to increase system efficiency, but they should also satisfy other desirable properties such as incentive compatibility and fairness. We analyze two reallocation mechanisms using a game-theoretic framework, and we develop simple, non-truthful strategies that airlines may use in these mechanisms. We show that the overall performance of the mechanisms can vary widely when gaming is introduced, and that enforcing fairness can have an large impact on the airlines’ best strategies. Jackie W. Baek, MIT, 77 Massachusetts Ave, Bldg E40-103, Cambridge, MA, 02139, United States, Hamsa Balakrishnan n TE35 North Bldg 224A Airline Competition and Coordination Sponsored: Aviation Applications Sponsored Session

n TE31 North Bldg 222A TSL Best Paper Session Sponsored: Transportation Science & Logistics Sponsored Session

Chair: Tom Van Woensel, Eindhoven University of Technology, Industrial Engineering, Den Dolech 2, Pav F08, Mb Eindhoven, NL5600, Netherlands 1 - TSL Best Paper Award Tom Van Woensel, Eindhoven University of Technology, Industrial Engineering, Den Dolech 2, Pav F08, Mb Eindhoven, NL5600, Netherlands The TSL Best Paper Award is given every year to an outstanding paper in the field of transportation science and logistics. The paper must have been published in a refereed journal and must present innovative approaches for solving complex problems in transportation and/or logistics, with an emphasis on operations research and quantitative methods. n TE32 North Bldg 222B Practice - Reverse Logistics/ Manufacturing I Contributed Session Chair: Debdatta Sinha Roy, Robert H. Smith School of Business, University of Maryland,College Park, MD, 20742, United States 1 - Evaluating Profitability of Remanufacturing Operations Akshay Mutha, University of Vermont, Grossman School of Business, 55 Colchester Avenue, Burlington, VT, 05405, United States, Saurabh Bansal, V. Daniel R. Guide We compare different methods for evaluating profitability of remanufacturing operations. We show the application of our model to current industry practices. 2 - Modelling Price Dynamics for New and Remanufactured Smartphones Supanan Phantratanamongkol, University of Birmingham, Birmingham, United Kingdom, Gu Pang In this presentation, we model price dynamics of new and remanufactured smartphones listed on eBay by using functional data analysis. More specifically, we explore how price velocity and acceleration evolve over product life cycles. We further investigate how the impact of different factors on the price dynamics may vary. Our results shed light on the price dynamics of buy-it-now live listings and their price determinants guiding both eBay sellers and buyers in making informed decisions. 3 - Optimization Models to Enable Job Rotation Schemes for Worker Safety Amir Mehdizadeh, Auburn University, Auburn, AL, United States, Alexander Vinel, Mark Schall, Richard Sesek, Sean Gallagher We consider the problem of employing job rotation schemes to improve worker safety in a manufacturing setting by combining optimization methods with novel modeling techniques developed in the occupational safety community. The work is based on a recently proposed fatigue-failure model for musculoskeletal disorders (MSD) risk evaluation. This leads to nontrivial nonlinear and nonconvex optimization problems which we solve through a combination of analytical and numeric tools. We conduct a realistic case study and conclude that while in some cases job rotation can lead to improvements, its effect is highly dependent on the composition of the job pool. 4 - A Stochastic Multi-echelon Multi-channel Network Problem Yuan Wang, National University of Singapore, Singapore,119260, Singapore, 119260, Singapore, Loo Hay Lee, Ek Peng Chew To effectively plan supply chain operations, it is meaningful to integrate strategic decisions on manufacturing capacity with short-term production and distribution decisions considering both demand and supply uncertainties. However, research on integrated supply chain network planning is rather limited in the literature. In this research, a stochastic multi-echelon multi-channel network model is proposed to solve the real supply chain network planning problem with consideration of uncertain demand forecast, production yield and raw material procurement simultaneously. The objective is to minimize overall operating cost through an extended planning horizon. 5 - Voting Mechanism Design in Randomized Strategic Social Choice Debdatta Sinha Roy, Robert H. Smith School of Business, University of Maryland, College Park, MD, 20742, United States, Debasis Mishra Consider a voting setup where each agent has an ordering over a set of alternatives. A randomized social choice function (RSCF) assigns probability distributions over the set of alternatives based on the orderings. A voting

391