2016 INFORMS Annual Meeting Program

TA86

INFORMS Nashville – 2016

2 - Managing Demand Uncertainty in Collaboration Mechanism Design for Carrier Alliances

3 - An Approximate Solution Approach For Blood Management When There Are Multiple Independent Sources Of Supply David C Novak, Associate Professor, University of Vermont, 310 Kalkin Hall, 55 Colchester Avenue, Burlington, VT, 05405-0157, United States, dnovak@bsad.uvm.edu, Marilyn T Lucas, S. Karti Puranam We present an approximation method to solve the infinite horizon, fixed lifetime perishable, inventory model with a lifetime of m > 2 periods, where there are two independent sources of supply. One source is blood ordered by the blood bank. The other source is blood that is randomly transferred from smaller, lower-usage hospitals in a regional blood exchange network to blood bank. We formulate a DP to solve the multi-period cost minimization problem and test our solution approach both theoretically and empirically. 4 - Improve Service Levels And Reduce Labor Cost Using Different Float Nursing Level Strategies Under Staff Absences In Hospitals Nurse shortage is one of major problems in Healthcare Systems. Float nurse strategy is an alternative solution method in hospitals to minimize nurse shortage issue. In this research, we investigate service level, staff absences and economic effects of nurse shortage to find a core regular unit nurse level with flexible float nursing strategy. Under different regular nurse core levels, optimal service and float nurse levels are determined. Performance of our model is also compared with different service level methods in literature. Case study results demonstrate our model provide higher service level and decrease labor cost comparing with current practice in a local hospital. 5 - Design Of Multi-stage Multi-provider Hybrid Appointment System For Patient Scheduling Under Uncertainty Sharan Srinivas, PhD Candidate and Research Assistant, Pennsylvania State University, 310 Leonhard Building, University Park, PA, 16802, United States, sus412@psu.edu, Arunachalam Ravindran Recent research focuses on designing hybrid appointment systems (HAS) for patient scheduling by combining open access and pre-booking scheduling methods. However, the multi-stage nature of patient flow, patient availability and uncertainties in outpatient clinics are rarely integrated in the design. We propose a deterministic model, and scenario based Monte Carlo approach to address this gap. The proposed approach aims to improve patient satisfaction and resource utilization by determining the percentage of appointments reserved for pre- booking and open access. A case study with real data from a Family Medicine clinic is used to show the feasibility of the proposed approach. Kamil Ciftci, Lehigh University, 200 West Packer Avenue, Bethlehem, PA, 18015, United States, kac208@lehigh.edu Chair: Ryan Choi, Assistant Professor of Marketing and SCM, Eastern Michigan University, 300 W. Michigan Ave., College of Business, Eastern Michigan University, Ypsilanti, MI, 48197, United States, jchoi20@emich.edu 1 - Investigating The Impact Of Social Influence On The Personalization-privacy Paradox: An Eye Tracking Study Thomas Frick, PhD Student, Rotterdam School of Management, Burgemeester Oudlaan 50, Rotterdam, 3062PA, Netherlands, frick@rsm.nl, Ting Li, Paul Pavlou Using consumers’ personal information to personalize ads does not only increase perceived ad relevance but also triggers consumer privacy concerns. To study this personalization-privacy paradox, we use eye-tracking technology and investigate how social influence affects consumers’ perceived ad relevance and privacy concerns. By objectively measuring visual attention, we obtain a rich understanding of how users affectively and cognitively process information and assess ads. Our results provide insights into the mediating role of attention within the personalization-privacy paradox. 2 - Non-contractual Customer Retention In Multichannel Settings Chun-Wei Chang, Assistant Professor, Governors State University, 1 University Parkway, University Park, IL, 60484-0975, United States, cchang@govst.edu We present a framework for estimating multichannel customer relationship dynamics in a non-contractual setting that flexibly allows for relationship revival and investigates the effects of different channel experiences and marketing communication on retention and profitability. We use a multi-segment, multivariate hidden Markov modeling framework to model three managerially relevant customer behaviors: purchase amount, purchase incidence, and channel choice. We uncover two latent relationship states that customers migrate to and from - an active state and an inactive state characterized by different levels of purchase frequency, responsiveness to marketing, and profitability. TA86 GIbson Board Room-Omni Marketing V Contributed Session

Yuhan Wang, University of California, Irvine, CA, wangyuhan1101@gmail.com, Luyi Gui, Ozlem Ergun

A carrier alliance refers to a cooperative among transportation companies that often collaborate via sharing service network capacities. In this paper, we consid- er a type of collaboration mechanism via capacity exchange prices that has been widely adopted in practice, and aim to provide a comprehensive analysis of its coordination effectiveness under demand uncertainty. In particular, we analyze the structure of service networks of sea cargo alliances in practice and develop a decomposition algorithm to not only much simplifies the problem but also enables a detailed analysis into the structure of a robust exchange prices and the capacity-demand properties of networks where such prices exist. 3 - Interdiction Learning-based Approaches To Combat Security Threats On Information Systems Forough Enayaty Ahangar, University of Arkansas, Fayetteville, AR, United States, fenayaty@email.uark.edu, Chase Rainwater We consider a information system connected across a network of servers. In part one of the talk, we solve an interdiction-based model to strategically determine how content is allocated amongst available servers so to minimize the impact of a denial of service attack. In part two of the talk, we provide an operational framework for identifying network threats across the chosen network structure via a learning-based framework. Roles of optimization within this framework are highlighted and the methodology is applied to network data taken from a national laboratories computer logs. 4 - Optimal Linepack Planning Models For Gas Transmission Network Trung Hieu Tran, Postdoctoral Research Fellow, The University of Warwick, Coventry, CV4 7AL, United Kingdom, t.h.tran@warwick.ac.uk, Simon French, Rhys Ashman, Edward Kent, Mark Hamling, Ben Dickel National Grid, the gas network operator in the UK, experiences challenges maintaining pressure and linepack (quantity of gas in network) limits due to the transient behaviour of customers in an open market. In this paper, 2 mixed- integer programming models are proposed for optimal linepack planning (i.e. considering the compressibility of natural gas in pipelines) to compensate for the fluctuation of gas flows. The first model minimizes total deviation between simulated & target linepack such that all demand is satisfied. The second model determines time and actions to minimize total cost for resolving linepack deficit. The efficiency of models has been validated in case studies at National Grid. Chair: Sharan Srinivas, PhD Candidate and Research Assistant, The Pennsylvania State University, 310 Leonhard Building, University Park, PA, 16802, United States, sus412@psu.edu 1 - Interaction Between Operational Efficiency And Doctor Incentives In Outpatient Services Guoming Lai, UT Austin, 1 University Station, B6500, Austin, TX, 78712, United States, laiguoming@gmail.com, Xiaofang Wang This paper studies the interaction between a doctor and a population of patients in a congested health service delivery system. The doctor’s prescribing decisions depend on her diagnostic/treatment ability, level of altruism and the institutional framework. Some patients are strategic and decide to see this doctor based on perceived quality, congestion and monetary costs. Within such a setting, we study the socially optimal decisions and provide policy insights. 2 - Equitable Nurse Scheduling By Goal Programming Esra Agca Aktunc, Assistant Professor, Kadir Has University, Kadir Has Caddesi Cibali, Istanbul, 34083, Turkey, esra.agca@khas.edu.tr Hospitals have to provide continuous service by employing the shift system and workers, mainly doctors and nurses, are required to work efficiently to avoid errors. Quality of healthcare services can be improved significantly if the nurse shifts are scheduled according to nurses’ preferences and by distributing the workload equitably. Schedules should also abide by hospital policies and workload requirements in each shift by assigning nurses with different skill sets. In this study, monthly nurse scheduling problem is modeled and solved by goal programming observing goals that represent nurse and patient satisfaction with fairness measures such as the number of night shifts and weekend shifts. TA79 Legends G- Omni Health Care, Modeling VX Contributed Session

261