2015 Informs Annual Meeting

WB40

INFORMS Philadelphia – 2015

WB39 39-Room 100, CC Channel Management and Pricing Cluster: Operations/Marketing Interface Invited Session

3 - Managing Access to Primary Care Facilities Sina Faridimehr, Wayne State University, 4815 Fourth St., Detroit, MI, United States of America, fb1562@wayne.edu, Ratna Babu Chinnam In this research, we propose an MDP model to improve timely access for patients while maintaining clinic capacity utilization in primary care facilities. The model leverages correlations between scheduling practice, panel size management and access performance. Results from testing the models at VA facilities are promising. 4 - Comparing Emergency Room Performances Before and After Initiating Full Capacity Protocol Suman Mallik, University of Kansas, 1300 Sunnyside Ave, Lawrence, KS, 66045, United States of America, suman@ku.edu, Mazhar Arikan, Lu Wang Using a data from a large teaching hospital we compare the emergency department operating performances before after initiation of the full capacity protocol (a set of rules designed to alleviate crowding. 5 - Why are Medical Device Connectivity Standards so Elusive?

Chair: Shuya Yin, University of California, Irvine, Merage School of Business, Irvine, United States of America, shuya.yin@uci.edu Co-Chair: Saibal Ray, Professor, McGill University, 1001 Sherbrooke Street West, Montreal, Canada, saibal.ray@mcgill.ca 1 - Strategic Value of Bogo Offers under Competition Sreekumar Bhaskaran, sbhaskar@mail.cox.smu.edu, Saibal Ray, Haresh Gurnani In the grocery and retail industries, firms routinely offer BOGO (buy one get one off) offers in which a consumer is able to obtain lower price on larger quantity purchases. We examine the a firm’s pricing decision for such BOGO bundles under competition. The effect of consumer heterogeneity and market characteristics on this decision is also considered. Shuya Yin, University of California, Irvine, Merage School of Business, Irvine, United States of America, shuya.yin@uci.edu, Saibal Ray, Houcai Shen, Wenju Niu, Mehmet Gumus Guaranteed buyback programs offer customers protection against price drops and encourage them to upgrade their products. This, in turn, motivates manufacturers to improve and innovate their products. These insurance contracts are often offered by third-parties. Our goal is to understand how such programs impact the preferences of the manufacturer and the customers. 3 - Analyzing the Entry of Big-box Retailers in an Emerging Market We consider the impact of the entry of a big-box retailer (BBR) in a market dominated by small, mom-and-pop retailers. The small retailers are characterized by local coverage of the market, whereas BBR provides services valued by all customers. Since both types of retailers obtain supplies from a common manufacturer, BBR’s entry affects the supply conditions. Our work thus highlights roles of direct competition as well as indirect supply side effect on small retailers and customers. 4 - Channel Contract Preferences under Dynamic Market Conditions Long Gao, Associate Professor of Operations and Supply Chain Mgt, University of California Riverside, 900 University Ave, Riverside, CA, 92507, United States of America, long.gao@ucr.edu Downstream retailers often have private information about consumer market conditions that may evolve over time. We study the long-term channel contracting problem under market fluctuations. We characterize the optimal contract, and show that incorporating market evolution is critical for contract design and execution. WB40 40- Room 101, CC Operations Management/Marketing Interface I Contributed Session Chair: Alejandro Lamas, Assistant Professor, NEOMA Business School, 1 Rue du MarÈchal Juin, Mont Saint Aignan Cedex, 76825, France, alejandro.lamas@neoma-bs.fr 1 - Competitive Time-Locked Free Trial Strategy 2 - Signing Up for Guaranteed Buyback Programs? Perspectives of Manufacturers and Customers Aditya Jain, Baruch College, New York City, New York, United States of America, Aditya_Jain@isb.edu, Saibal Ray, Mehmet Gumus

John Zaleski, Chief Informatics Officer, Nuvon, Inc., 4801 S. Broad Street, Suite 120, Philadelphia, PA, 19112, United States of America, jzaleski@nuvon.com

Medical devices still remain highly proprietary in terms of interoperability. Health Level Seven (HL7), as a healthcare information standard, only works when medical devices can export data in this common format. Gaps remain between the proprietary, manufacturer-specific language of many devices and the HL7 messaging format. Here we explore approaches for standardizing proprietary equipment around HL7 and related messaging languages and how lack of interoperability impacts patient care.

WB38 38-Room 415, Marriott Bayesian Approach II Contributed Session

Chair: Ray Fung, Self-Employed, 10 Soden Street, #16, Cambridge, MA, 02139, United States of America, raymondfung@gmail.com 1 - Enterprise Personalized Learning At Scale Ashish Jagmohan, IBM Research, 1101 Kitchawan Road, Yorktown Heights, United States of America, ashishja@us.ibm.com, Wesley Gifford, Anshul Sheopuri, Yi-min Chee, Noi Sukaviriya, John Ambrose, Laura Rexford, Sue Rodeman We address the problem of facilitating skill development in enterprises by devising personalized learning paths. Traditional methods of manual curation are incapable of scaling to meet the needs of growing user bases and content-libraries. The proposed system uses big-data cognitive technology to reason about large-scale user behavior and content characteristics. We will discuss algorithmic data mining and Bayesian techniques to identify learning sequences best suited for each user’s goals. 2 - Identifying Key Rule-based Subgroups for Driver Or Graphical Models via Modified Decision Tree Logic Michael Egner, Senior Vice President, Ipsos, 10567 Jefferson Blvd., Culver City, CA, 90232, United States of America, mike.egner@ipsos.com, Andrew Christianson, Richard Timpone Modelers often wish to understand the role of moderator variables. In some cases, such as graphical models, the typical solution (adding interaction terms) can fall short. Building on previous research modifying decision trees to split on the strength of bivariate relationships, this study explores modifying tree logic to maximize differences in association matrices, such that researchers can obtain practical, rule-based splits for generating maximally-different drivers or graphical models. 3 - Causal vs. Correlational Analysis using Bayesian Networks Ray Fung, Self-Employed, 10 Soden Street, #16, Cambridge, MA, 02139, United States of America, raymondfung@gmail.com I show how Bayesian Networks can be utilized not only to differentiate correlation and causation in an intuitive manner but also how to illuminate difficult-to-understand scenarios such as Simpson’s Paradox, the Ecological Fallacy, and the Low Birthweight Paradox. I also show how concepts such as instrumental variables, LATE, overidentification tests, natural experiments, ITT, block randomization, mediation tests, colliders, and measurement error can be easily illustrated.

Hai-ping Wang, Xiëan Jiaotong University, School of Management, Xi’an, China, whp1989@stu.xjtu.edu.cn, Jun Lin, Shu-lin Liu

Offering time-locked free trials has been a common practice in the software industry to reduce consumers’ uncertainty about product quality. This paper develops a game-theoretic model to determine the optimal time-locked free trial strategy in a duopoly market. Allowing consumers to have very different, and even opposite experiences after trial, this paper analyzes the impact of consumer learning heterogeneity on the equilibrium outcome. Several new and important insights are provided.

413