Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

WB26

3 - Social Determinants of Health Information Technology Enabled Patient Provider Engagement among Patients with Multimorbidity Ajit Appari, Worcester Polytechnic Institute, Worester, MA, 77030, United States, Meghan Hufstader-Gabriel Our understanding of health information technology (IT) enabled patient- provider engagement by patients with multimorbidity is limited despite growing trends of health IT diffusion. We analyzed 2017-Health Information National Trends Survey data on American adults (18+yrs old) using weighted ordered logistic regression to evaluate association of health IT enabled engagement (Low, moderate, High) with demographic and socioeconomic factors for Low (2-3 conditions) and High (4+ conditions) multimorbidity groups separately. Out study did not find association of health IT enabled engagement with any covariate, except older patients more likely to engage. 4 - Online Demand Driven Car Sharing Rebalancing Xiaopeng Li, Dr., University of South Florida, Tampa, FL, 33620, United States, Dongfang Zhao This study proposes an online car-sharing rebalancing model to deal with the car sharing fleet management problem using reinforcement learning. We develop a multi-agent reinforcement learning framework using a deep Q-learning algorithm. The goal of the algorithm is to maximize the total profit of the platform by repositioning available cars to the locations with outstanding demand-supply gaps. This model does not make any assumption on the demand and is completely driven by spatially distributed demand data history. We show significant improvements of the proposed framework over state-of-the-art approaches through extensive empirical studies. 5 - Service Delivery in the Sharing Economy: Platform Assumptions and Technology Affordances for Prosumer Service Interactions Anita D. Bhappu, University of California Merced, 5200 N. Lake We develop a framework for understanding the nature of service delivery in the sharing economy by applying the literature on service science to three exemplar ridesharing platforms. To facilitate and support our analysis, we critically examine and discuss these platforms’ implicit assumptions about service interactions and prosumer motivations as evidenced by their designed technology features. We also assess these platforms’ technology affordances for service governance and prosumer trust. n WB26 North Bldg 132A Stochastic Models for Service Operations Sponsored: Service Science Sponsored Session Chair: Xu Sun, Columbia University, New York City, NY, 10027, United States This paper establishes tight upper bounds for the mean stead-state and transient waiting times in the GI/GI/1 queue given the first two moments of the inter- arrival times and service-time distributions. We apply the theory for the classical moment problem to show, for distributions with support on bounded intervals, that the bounds are attained at distributions with support on at most three points. For distributions with support on the unbounded positive real line, we show that the bounds are not attained directly, but are obtained asymptotically. We propose a simple approximation formula and provide a numerical comparison of the approximations and bounds. 2 - Optimal Charging Control in an Electric Vehicle Battery Swap Station Xu Sun, Columbia University, New York City, NY, 10027, United States, Bo Sun, Ward Whitt We model an electric vehicle battery swap station as a closed-loop production- inventory system facing time-varying demand for battery swap and time-varying prices for recharging batteries. The objective is to find an optimal charging policy that best trade off the charging cost and the congestion. We first formulate a Markov decision process (MDO) and then develop a fluid model approximation for the (MDP). We gain important managerial insights by solving the fluid model optimization. Road, SSM 233, Merced, CA, 95343, United States, Melanie Lisa Yeo, Ann Kovalchick, Tea Lempiala 1 - Extremal Single-Server Queues Given Two Moments Yan Chen, Columbia University, New York, NY, 10027, United States, Ward Whitt

n WB24 North Bldg 131B Finance - Theory & Empirics Contributed Session Chair: Peng Xie, California State University, Hayward, CA, 94542, United States 1 - Politics and Inventory: Chief Executive Officer – White House Political Ideology Alignment and Inventory Management Mazhar Arikan, University of Kansas, Lawrence, KS, 66045, United States, Yi Tan, Yaoyi Xi We empirically examine whether and how CEO political ideology affects inventory strategies. Using a sample of 3,461 distinct U.S. domestic publicly traded firms, we find that firms tend to hold more inventories during the periods when their CEOs’ personal political ideology (measured by donations to political parties) is in accordance with the President of the U.S. We perform further analyses on firm financial performance and economic policy uncertainty to investigate the underlying mechanisms for this finding. Results indicate that CEO- President political ideology alignment does not provide an information channel through which CEOs can exploit additional value-related information. 2 - Distance-based Portfolio Combining Algorithm to Predict in Sample Optimal Portfolio Hongseon Kim, Yonsei University, ShinChonDong, Daewoo Hall 610, Seoul, Korea, Republic of We provide portfolio combination algorithms for reducing Euclidean distance between in-sample portfolio and out-of-sample portfolio. These algorithms combine various out-of-sample portfolio depending on the length of the Euclidean distance and tracking signal. We conducted an empirical experiment using 21 data sets. We find that combined portfolio which we propose often has a short Euclidean norm and higher Sharpe ratio than other benchmark portfolios. 3 - The Effect of Initial Coin Offering Uniqueness on its Valuation and Trading Status Peng Xie, Assistant Professor, California State University, Hayward, CA, 94542, United States As the development of the cryptocurrency ecosystem, a new way to raise funds for new cryptocurrency ventures called Initial Coin Offering (ICO) becomes popular. Using data from more than 1,000 ICO campaigns from 2017 to 2018, we empirically show that the uniqueness of the ICO ideas is positively correlated with the cumulative returns and the probability of ongoing trading in cryptocurrency exchanges. n WB25 North Bldg 131C Service Sharing Sponsored: Service Science Sponsored Session Chair: Anita D. Bhappu, University of California-Merced, Merced, CA, 95343, United States 1 - Bonus Scheme Design in Ride-sharing Economy under Competition Puping Jiang, PhD, Washington University in St.Louis, Olin Business School, Campus Box 1156, St.Louis, MO, 63130-4899, United States, Fuqiang Zhang, Kaitlin Daniels Nowadays in the ride-sharing economy, one phenomenon takes our attention is the platform’s “convex”-like bonus policy towards drivers. In this research we mainly address two questions: First, what is the main driving force behind such bonus schemes? Second, why does the bonus scheme have the “convex”-like form? This is the first research in ride-sharing economy as far as we know looking at the competition between two ride-sharing platforms. 2 - What Types of Drivers Do Our Society Need? The Effect of Drivers’ Refusing Behavior Xiaojing Feng, Shanghai Jiao Tong University, Minneapolis, MN, 55454, United States, Ying Rong, Tony H. Cui Although the empirical evidence shows that the driver’s strategic refusing behavior is one of the key factors that contribute to high Incomes, it harms the feelings of the refused passengers, which may cause passengers’ complaints. We propose an analytical model to capture the refusing behavior of different types of drivers, depict the game dynamics between drivers, and try to explore the effect of the refusing behavior on the social welfare under different regulations.

449