INFORMS 2021 Program Book

INFORMS Anaheim 2021

WD21

4 - Analyzing the Language Used in Fake News Related to Covid-19 Alireza Farnoush, Auburn University, Auburn, AL, United States, Ashish Gupta, Gregory Purdy The flood of fake news in a situation of a pandemic such as COVID-19 which affected many people can result in serious effects on societies. In this study, we aim to analyze the language used in fake associated with COVID-19. We provided the theoretical foundation needed to develop a better understanding of the science of fake news. We also raised different hypotheses and test them by applying the text analytics method and component-based SEM. Our finding indicates that fake news in COVID-19 has more uncertainty, less complexity, more negative affect, less specificity, less information quantity, less diversity, more expressivity words, and more cognitive process. WD21 CC Room 204A In Person: Sustainability and Emerging Technologies General Session Chair: Saed Alizamir, Yale University, New Haven, CT, 6520, United States Chair: Michael Blair, New Haven, CT, 06510-1008, United States 1 - Two-sided Benefits of Price Transparency in Informal Supply Chains Yuan Shi, Massachusetts Institute of Technology, Cambridge, MA, United States, Joann de Zegher, Irene Yuan Lo This paper develops a new Hotelling model of price search for the welfare impact of price transparency in informal supply chains. The model incorporates the price- setters’ operations under downstream contractual obligations and informal business relationships. We show that under demand asymmetry and costly supply uncertainty, a moderate increase in price transparency leads to a strong Pareto improvement in a competitive duopoly market. This effect persists under price collusion. Our findings contrast with the typical assumption that increased price transparency leads to one-sided benefits at the cost of the other side and inform the design of information platforms in informal supply chains. 2 - When the Wind of Change Blows, Build Batteries? Optimum Renewable Generation and Energy Storage Investments. Christian Kaps, PA, United States, Simone Marinesi, Serguie Netessine Renewables have become the cheapest energy source in most of the world, but their generation remains variable and difficult to predict. Recent technological advances have rendered large-scale electricity storage economically viable, thus mitigating the renewable intermittency issue. However, it is not yet well- understood how to jointly determine optimal capacity for their generation and storage. Our work aims to shed light on this question by developing a two- product newsvendor model of a utility’s strategic capacity investment in renewable generation and storage to match demand with supply, while using fossil-fuel backup, if needed. 3 - Leveraging Smart Thermostat Data to understand the Impact of Climate Change on Residential Energy Consumption Michael Blair, Yale University, New Haven, CT, 06510-1008, United States, Saed Alizamir, Shouqiang Wang In this work, we empirically analyze a rich micro-level thermostat data set provided by a large smart thermostat manufacturer. Our analysis reveals that households differ significantly in how they utilize their thermostats. This is partially driven by heterogeneity in daily occupancy schedules, but more importantly, many households do not utilize the essential features of their thermostat. This suboptimal behavior leads to increases in consumption that are magnified on extremely hot or cold days. We also see that even small changes to our climate will have a profound impact on consumption, and this impact is correlated with how a household uses their thermostat. This highlights the potential value of smart technologies, but also the importance of using these products properly.

WD22 CC Room 204B In Person: Health Care, Modeling and Optimization I Contributed Session Chair: Pritom Kumar Mondal, Texas Tech University, Lubbock, TX, 79407-2622, United States 1 - The Provider Network Selection Problem in Healthcare Markets Amin Hosseininasab, Warrington College of Business, University of Florida, Gainesville, FL, United States, Willem-Jan van Hoeve, Sridhar R. Tayur Provider network selection is a central problem faced by the healthcare insurance industry. A provider network consists of healthcare providers that are contracted by an insurer in order to provide healthcare services at discounted prices to insured patients. The problem involves insurance plan and network design to target patients under competition. We develop a novel methodology for optimal provider network selection, and show that our approach improves over methods used in the literature and a real-world insurer on test instances. We then use our methodology to analyze a number of insurance policies and mandates in terms of their effects on social welfare, equity, profits, and expenditures. 2 - A Continuous Scoring Model for Fair Liver Transplant Allocation Subramanian Raghavan, University of Maryland-College Park, College Park, MD, United States, Shubham Akshat The United States (U.S.) Department of Health and Human Services is interested in increasing geographical equity in access to liver transplant. We develop a novel analytical method to design heterogeneous scoring functions for continuous scoring policy in the deceased donor liver transplantation that equalizes supply to demand ratios across the transplant centers. The framework is general enough to be applied to other organs as well. 3 - Multi-Hospital Surgical Block Scheduling Candace Arai Yano, University of California-Berkeley, Berkeley, CA, United States, Alexandra M. Newman, Vishrut Rana Mergers and organic growth of healthcare systems have led to systems with multiple hospitals in the same general vicinity. These healthcare systems are considering whether to consolidate certain surgical procedures in a limited number of locations, thereby better justifying the purchase of advanced surgical equipment such as robots via higher utilization of this equipment. We address a multi-hospital surgical block scheduling problem that allows for such consolidation while accounting for nursing costs, capacity in downstream hospital wards and numerous other realistic factors while also limiting the assignment of surgical patients to distant hospitals for simpler surgical procedures. 4 - Optimizing Equitable Access to Emergency Care in San Francisco Robert Newton, PhD Student, Pennsylvania State University, University Park, PA, United States San Francisco’s “Blueprint” plans to re-open post-pandemic include a metric of healthcare equity but look solely at number of insured adults. Previous work suggested inequity in the distribution the city’s emergency rooms, leaving lower income areas underserviced. We consider two models to optimize ER locations— Set Covering (SCP) and Maximal Coverage (MCLP). The SCP suggests seven ERs can service the city with four new locations not controlling for capacity. The MCLP better accounts for existing ERs and proposes a new ER in ZIP code 94112 to maximize coverage of underserviced areas—tripling the probability 911 calls from low-income ZIP codes are within the threshold distance to an emergency room. 5 - Capacity Planning in a Psychiatric Hospital Using Mixed Integer Linear Programming Pritom Kumar Mondal, PhD Student, Texas Tech University, Lubbock, TX, United States, Bryan A. Norman Severely ill psychiatric patients from outpatient mental health settings are referred to a psychiatric hospital to potentially receive treatment. However, due to not having sufficient mental health therapists, patients do not always receive therapy treatments in a timely manner, which hinders their quality of care and health status. A mixed integer linear program has been developed to determine the number of mental health therapists required to ensure patients get all necessary therapies at the appropriate time.

170