Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

MA40

2 - Capabilities-based Planning for Emergency Preparedness Cameron MacKenzie, Iowa State University, Ames, IA, United States, Rohit Suresh The Federal Emergency Management Agency’s (FEMA) National Preparedness System identifies 32 core capabilities for emergency preparedness. State and local agencies must assess how well their organizations are performing in each of those 32 capabilities. However, measuring these capabilities is very difficult and understanding how these capabilities relate to each other is also very challenging. We apply the principles of value-focused thinking and Bayesian belief networks to help states to conduct capabilities-based planning for emergency preparedness. 3 - Detection of Fake Social Media Postings Following Soft-target Terror Attacks: A Signal Detection ROC Analysis Richard S. John, University of Southern California, Dept of Psychology MC 1061, Los Angeles, CA, 90089-1061, United States, Mengtian Zhao We investigated the ability of individuals to detect false information from actual posts on social media platforms. In 3 experiments manipulating the base rates for false information and the penalty for incorrect classification, US participants viewed Tweets posted shortly after terrorist attacks on soft targets in the US and Europe, and judged the authenticity of information for each post. Using a signal detection framework, participants performed only slightly better than chance, consistent with previous research on lie detection. Respondents were relatively insensitive to manipulation of base-rates and relative error penalties. 4 - Analyzing, Modeling, and Detecting the Spread of Misinformation on Social Media during Natural Disasters Kyle Hunt, Student, University at Buffalo, Buffalo, NY, 14228, United States, Bairong Wang, Jun Zhuang Our studies investigate the spread of misinformation on social media through big data analysis and mathematical modeling. We implement techniques such as network analysis, sentiment analysis, predictive modeling, and decision modeling to understand the dissemination features of rumor spreading and rumor debunking information across Twitter’s network. Results from data analysis show that official debunks from government accounts were very effective in combating the spread of rumors, which can offer valuable insight on the optimal use of such accounts. Results from current studies will show the attitudes and behaviors of Twitter users during times of rumor spreading on Twitter. 5 - Prepare or Wait? The Marine Forces Reserve Hurricane Decision Simulator Eva D. Regnier, Naval Postgraduate School, 555 Dyer Road, Bldg. 330, Room 287, Monterey, CA, 93943, United States, Cameron MacKenzie After 2012’s Hurricane Ivan, the U.S. Marine Forces Reserve (MFR) asked for help with a challenging decision-whether and when to incur up to $10M in direct cost to prepare for a hurricane that may or may not affect their operations. We worked with Marines and other local officials to understand and model their decision processes and used data analytics to build a storm-simulation model. The result is an online training tool that the MFR has used since 2016. The simulator lets individual users experience decades’ worth of storms in a few hours. In team exercises, MFR reports that the HDS training provides a more realistic experience than hand-generated scenarios used previously. n MA43 North Bldg 227B Energy and Climate II Emerging Topic: Energy and Climate Emerging Topic Session Chair: Valerie Thomas, Georgia Tech, GA, United States 1 - Quantification of Technological Uncertainty in Evaluating the Feasibility of Meeting Emission Targets Dong Gu Choi, Pohang University of Science and Technology (POSTECH), 77 Chengam-Ro, Nam-Gu, Pohang, Gyeongbuk, 37673, Korea, Republic of, Hansung Kim Recent studies have considered uncertainty for GHG reductions by using the stochastic programming model for energy systems. The models require values of input parameters in the form of a scenario tree to handle uncertainty. However, only few studies have used quantitative methods to generate scenarios of the future values of technological input parameters. We show that the scenarios of the future values of the technological input parameter for energy system models can also be generated by a quantitative methodology. In addition, we introduce how to quantify the risk for the feasibility of meeting GHG emission targets by incorporating the technological uncertainty.

n MA40 North Bldg 226B Joint Session APS/RMP: Revenue Management and Pricing in Service Systems Sponsored: Applied Probability Sponsored Session Chair: Yasar Levent Kocaga, Sy Syms School of Business, New York, NY, 10033, United States 1 - Operational Perils and Benefits of Free Trials in Large Scale Service Systems Yasar Levent Kocaga, Sy Syms School of Business, Belfer Hall Room # 403/A, 2495 Amsterdam Ave, New York, NY, 10033, United States, Chihoon Lee We consider a firm that is serving price and delay sensitive customers and that has the option of offering a free trial service to a new market of customers. We first show that it is optimal for the revenue maximizing firm to operate in the QED heavy traffic regime and provide tractable and accurate expressions for the optimal price and revenue. Then, we use these expressions to identify conditions John Hasenbein, University of Texas-Austin, 1 University Station Stop C2200, Department of Mechanical Engineering, Austin, TX, 78712-0292, United States, Chengcheng Liu We study Naor’s model in which the arrival rate is uncertain and customers are heterogeneous in either their value of service or holding cost per unit time, or both. Even characterizing the stability of such systems is non-trivial. We focus on the case when the queue length is observable to the customers, and allow a social optimizer or revenue maximizer to manage the system. 3 - Managing Service Systems via Disguised Queues: The Role of Customers’ Retaliatory Behavior Eren Basar Cil, University of Oregon, 1208 University of Oregon, Lindquist College of Business, Eugene, OR, 97403-1208, United States This paper studies firms that can partially disguise their waiting lines to influence customer demand. We investigate the impacts of customers’ retaliatory behavior on the firm’s queue hiding decisions and profits. We show that firms can significantly benefit from disguised queues if customers retaliate to queue disguising tactics whereas any disguised queues hurt firms when they face non- retaliating customers. 4 - Asymptotic Analysis of Multi Queue Service Systems with Dynamic Customer Choice Yichuan Ding, University of British Columbia, University of British Columbia, 6333 Larkin Drive, Vancouver, BC, V6T 1C3, Canada, Mahesh Nagarajan, Zhe Zhang This paper introduces a multiclass queuing model to study a stochastic service system with multiple servers and dynamic consumer choice. We consider service providers of heterogeneous quality distributed at different locations with heterogeneous customers arriving randomly to the system. An arriving customer chooses a service provider to obtain service. We prove that the fluid limit process has a unique equilibrium, and the diffusion limit process is a reflected multi- dimensional Ornstein-Uhlenbeck process centered at that equilibrium. under which offering free trials is beneficial to the firm. 2 - Naor’s Model with Uncertain Arrival Rates and Heterogeneous Customers

n MA41 North Bldg 226C Decision Analysis, Game Theory, Big Data, and

Disaster Management Sponsored: Decision Analysis Sponsored Session

Chair: Jun Zhuang, University at Buffalo, Buffalo, NY 1 - Game-theoretic Modeling of Pre-disaster Relocation

Vicki Bier, University of Wisconsin-Madison, 1513 University Avenue, Room 3234, Madison, WI, 53706, United States, Hongru Du, Yuqun Zhou This talk uses game theory in conjunction with standard exponential discounting to explore strategies by which governments might encourage pre-disaster relocation by residents living in high-risk areas. We find that offering a subsidy (e.g., a partial buyout) can frequently be effective, if government has a lower discount rate than residents. We also present extensions to our model, exploring the use of a fixed annual benefit after relocation (instead of a one-time subsidy), and hyperbolic instead of exponential discounting.

134