2015 Informs Annual Meeting

MA33

INFORMS Philadelphia – 2015

MA31 31-Room 408, Marriott Retail Analytics Sponsor: Data Mining Sponsored Session

relies on an original modeling architecture that integrates an Integer Programming scheduling model, a Dynamic Programming operational model and a Stochastic Queuing Model of congestion. Results suggest that operating enhancements and limited, targeted scheduling adjustments can significantly reduce delays at busy airports. 3 - Design and Analysis of Matching and Auction Markets Daniela Saban, Stanford University, 655 Knight Way, Stanford, CA, United States of America, dsaban@stanford.edu Auctions and matching mechanisms have become an increasingly important tool for planners to allocate scarce resources among competing individuals or firms. This thesis addresses several questions that arise when designing and analyzing such markets. For example, we design auctions to construct catalogs of goods for government use, and matching mechanisms that can potentially be used to handle appeals in the public school assignments of thousands of incoming high school students. MA33 33-Room 410, Marriott Policymaking in Public Health Sponsor: Health Applications Sponsored Session Chair: Ben Johnson, Georgia Institute of Technology North Av, Atlanta, GA, 30332, United States of America, benjohnson@gatech.edu 1 - HIV and STIS Among Young MSM and the Operational Issues of Expanding Testing Benjamin Armbruster, Northwestern University, 2514 Sheridan Rd, Evanston, IL, United States of America, armbrusterb@gmail.com We discuss the health policy conclusions one can draw from a detailed and validated agent-based network simulation model of HIV, gonorrhea, and chlamydia spread among young men who have sex with men (YMSM) in Chicago. We focus on racial disparities and the operational issues of expanding HIV and STI testing such as combined testing, cost-effectiveness, roll-out speed, and uptake behavior. 2 - Data Driven Approach to Bundled Payments Margret Bjarnadottir, Assistant Professor of Management Science and Statistics, Robert H. Smith School of Business, University of Maryland, 4324 Van Munching Hall, College Park, MD, 20742, Mike Carter, University of Toronto, Mechanical & Industrial Engineering, 5 King’s College Rd., Toronto, ON, M5S 3G8, Canada, mike.carter@utoronto.ca, Anna Graber, Vedat Verter Many countries experience disparities in the distribution of health professionals. There is evidence that trainees from a rural background are more likely to choose to practice in rural areas. Our proposed optimization model incorporates interests of two main stakeholders in the system, namely the regulator and the health professionals, and provides the optimal training locations and required background of trainees in each location. 4 - Evaluating Policy and Network Interventions to Improve Dental Accessibility and Availability for Children Ben Johnson, Georgia Institute of Technology, Georgia Institute of Technology North Av, Atlanta, GA, 30332, United States of America, benjohnson@gatech.edu, Nicoleta Serban, Paul Griffin, Susan Griffin We develop an intervention optimization model to match supply and need for pediatric dental care in Georgia under different dental care policies and provider networks. The model is used to assess the trade-off between efficiency (expenditure) and equity (systematic variations in accessibility and availability) under different policies for managing decision making and outcomes. Network and policy interventions are then determined to provide optimal improvements in dental access. United States of America, margret@rhsmith.umd.edu, Wenchang Zhang, Ruben Proano, David Anderson, Renata Konrad Healthcare reimbursement is at the forefront of healthcare reform debates. Bundled payments reimburse a single amount for an episode of care and have been proposed as a practical and promising reimbursement alternative to the current fee-for-service system. In this talk we investigate data driven methods to automatically define sets of services constituting episode of care. 3 - Improving the Equity of Access to Primary Care

Chair: Matthew Lanham, Doctoral Candidate, Virginia Tech, Dept of Business Information Technology, Blacksburg, VA, 24061, United States of America, malanham@gmail.com 1 - Assortment Planning for Consumers Learning their Tastes

Canan Ulu, Assistant Professor, Georgetown University, McDonough School of Business, Washington, DC, 20057, United States of America, Canan.Ulu@georgetown.edu, Dorothee Honhon

We study a firm that offers novel products to consumers who do not have set preferences. Consumers try different products to learn which product suits their tastes better. 2 - Optimal Stocking Decisions in a Multi-channel Retail Environment Nevin Mutlu, PhD Candidate, Virginia Tech, Department of Industrial and Systems Eng, Blacksburg, VA, 24061, United States of America, nmutlu@vt.edu, Ebru Bish, Erick Wikum As traditional brick-and-mortar retailers are expanding their sales channels to online and mobile channels, consumer adoption rates of these emerging channels is increasing over time. We develop a novel, dynamic demand model, and integrate it within an optimization model to understand the implications of this dynamic environment on the retailers’ optimal stocking decisions considering different products and different market settings. 3 - Inventory Mirroring in a Heterogeneous Network Inventory mirroring determines how many fulfillment centers (FC) and where each stock-keeping unit (SKU) should be stocked. Optimizing inventory mirroring is necessary when the FCs have SKU count limits. We propose an approximate inventory mirroring algorithm for a heterogeneous network, where the fulfillment centers (FC) have different capacities and SKU type eligibilities. We present analysis results of the output to validate the effectiveness of algorithm. 4 - Parameter Estimation Procedures for a Hierarchical Assortment Planning Decision Matthew Lanham, Doctoral Candidate, Virginia Tech, Dept of Business Information Technology, Blacksburg, VA, 24061, United States of America, malanham@gmail.com, Ralph Badinelli Estimating a consumer’s propensity to purchase a product as well as their substitution behavior are critical parameters to a retailer’s assortment decision. We investigate the methodologies used to understand consumer demand, substitution behavior, and formulate a novel approach that could be used strategically in a hierarchical assortment planning decision model. George B. Dantzig Dissertation Cluster: George B. Dantzig Dissertation Invited Session Chair: Nils Rudi, INSEAD, 1 Ayer Rajah Avenue, Singapore, 138676, Singapore, nils.rudi@insead.edu 1 - Robust Optimal Control for Medical Treatment Decisions Yuanhui Zhang, NC State, Raleigh, NC, United States of America, yuanhui.zhang@gmail.com In this dissertation, we develop a new data-driven robust stochastic optimization model for optimizing medical treatment decisions. We present computationally efficient methods for solving this model and theoretical analysis of the optimal policies. We illustrate the application of this model for optimizing treatment decisions for patients with type 2 diabetes and show that robust optimal policies could potentially provide guidance for clinicians and policy makers to make treatment decisions. 2 - Integrated Allocation and Utilization of Airport Capacity to Mitigate Air Traffic Congestion Alexandre Jacquillat, PhD Candidate, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Building E40-240, Cambridge, MA, 02116, United States of America, alexandre.jacquillat@gmail.com This thesis jointly optimizes airport operating procedures at the tactical level and flight scheduling interventions at the strategic level for congestion mitigation. It Zhiwei (Tony) Qin, Staff Data Scientist, Walmart Labs, 850 Cherry Ave, San Bruno, CA, 94066, United States of America, TQin@walmartlabs.com MA32 32-Room 409, Marriott

155