Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

MC71

4 - Xpress Mosel: Using Alternative Solvers and Trends in Modeling Susanne Heipcke, FICO, 181 Metro Dr, Ste 700, San Jose, CA, 92130, United States The modeling and solving environment Xpress Mosel has recently been turned into free software, simultaneously opening the Mosel language to alternative Mathematical Programming solvers. We give an overview of these solver interfaces and their use. The easy deployment of Mosel models as web applications without programming by adding configuration information to the model has raised demand for interconnection with analytics tools, such as Python. Another trend, the increasing use of Mosel for non-optimization tasks, has triggered a revision of the language, introducing a host of new programming- related features. n MC71 West Bldg 106C Joint Session ICS/Practice Curated: Retail Layout and Shelf Space Analytics Sponsored: Computing Sponsored Session Chair: Ahmed Ghoniem, University of Massachusetts Amherst, Amherst, MA, 01003, United States 1 - Discrete Event Simulation Model for Purchases at Temporary Displays in a Grocery Store Alice E. Smith, Auburn University, 3301 Shelby Center, Dept of Industrial/Sys Engineering, Auburn, AL, 36849, United States, Ilknur Uludag We develop a discrete event simulation model that evaluates impulse purchase likelihoods of product categories at temporary displays in grocery stores considering cross-category sales and customer paths. End-of aisles and islands are temporary displays for promotional, new, and high impulse products. In the simulation model, a customer shopping trip is created based on the number of purchased product categories, and a modified shortest path of the purchased products. The shopping basket data from a major grocery store chain, the impulse purchase likelihoods of product categories, and the permanent displays of product Bacel Maddah, American University of Beirut, Bechtel 519A, P.O. Box 11-0236 Riad El Solh, Beirut, 1107 2020, Lebanon, Ahmed Ghoniem, Tulay Flamand We investigate how the allocation of product categories can be optimized in a fashion that guides in-store traffic and stimulates impulse buying. We develop a predictive regression model that estimates traffic densities at a shelf as a function of the shelf space allocation and the location of the shelf in the store. This traffic model is embedded within a mixed-integer nonlinear program that determines the location and the shelf space of all product categories in the store in a way that maximizes impulse buying. To overcome the computational challenges, we utilize a variable neighborhood search. We apply our analytics framework to a grocery store in Beirut with encouraging results. 3 - Joint Consideration of Rack Layout and Product Allocation to Maximize Retail Impulse Purchase Pratik J. Parikh, Wright State University, 3640 Colonel Gelnn Highway, 207 Russ Engineering Center, Dayton, OH, 45435-0001, United States, Bradley R. Guthrie In a retail store, the configuration of racks and the placement of products on it affect what shoppers see during their shopping trip, and the eventual decision to buy. While research has shown that staple products (e.g., bread, milk) will be purchased no matter their location, the location of impulse products (e.g., candy, novelties) has a significant impact on their sales. We extend our prior research on quantifying exposure by proposing an optimization approach that determines the optimal rack orientation and curvature to maximize expected impulse profit across a variety of product location-allocation decisions. 4 - Hierarchical Optimization for Retail Store Layout Design Tulay Flamand, Colorado School of Mines, Division of Economics and Business, Engineering Hall 816 15th Street, Room 313, Golden, CO, 80401, United States, Ahmed Ghoniem, Bacel Maddah We address a layout design problem for a retail department store by proposing a hierarchical approach. First, we formulate a mixed-integer programming model, which allocates departments to different areas of the store in a way that jointly maximizes the customers’ shopping convenience and impulse buying. Using the prescribed allocation, we then propose a predictive model to form a main walkway around each department based on its traffic. Finally, we formulate a second optimization model that determines the specific location of each department category in its assigned area by considering the interplay between the category allocation, in-store traffic and impulse buying. categories in a grid store layout are analyzed by the simulation. 2 - Analytics for Store-wide Shelf Space Management

5 - The Price-setting Newsvendor with Poisson Demand Anna-Lena Sachs, University of Cologne, Albertus-Magnus-Platz, Cologne, 50923, Germany, Benedikt Schulte We study the price-setting newsvendor with Poisson demand. We develop an analytical solution approach that covers a broad class of demand curves, including both linear and iso-elastic demand curves, explain how to apply our approach to more general demand functions via piece-wise linear approximation, and develop analytical and numerical insights. We characterize the behavior of the optimal price and analyze the performance gap of different price-setting heuristics. We observe instances in which a significant share of profits would be lost if the discrete nature of demand were not modeled explicitly. n MC72 West Bldg 211A KINFORMS Sponsored: INFORMS Special Sessions Sponsored Session Chair: Chang Won Lee, Hanyang University, School of Business, 17 Haengdang Dong, Seoul, 133-791, Korea, Republic of 1 - An Event Study of Servitization: Focusing on Korean Companies Chongwoo Park, Korea University Business School, Seoul, Korea, Republic of, Hosun Rhim Korean economy has been grown up based on the solid manufacturing industries. As the economy gets mature servitization is regarded as a breakthrough. We examine impact of servitization to the shareholder value by measuring the stock market reaction associated with servitization announcement. 2 - A WoS-based Literature Review on Newsvendor Models Sungyong Choi, Yonsei University, College of Government and Business, 1 Yoneseidae-gil, Wonju, 26493, Korea, Republic of The newsvendor model has been actively studied in the field of management science and operations management as a stochastic inventory optimization model. Since the seminal work of Arrow et al. (1951), some variations of the newsvendor models have also been studied in literature until now. Then, using WoS (Web of Science), we will review the previous studies on the newsvendor models focusing on some well-known review papers and try to find further research directions. 3 - Forecasting Aviation Services Delay using MIT Lincoln Lab and EuroControl’s Estimation Criteria This study is to forecast aviation services delay by aviation weather. This study developed a forecasting model of the delayed cost of aviation service caused by aviation weather. We use the delay reduction model by MIT Lincoln Lab and EuroControl’s estimation criteria for delay cost 4 - A DEA-AR Model with Weight Bounds Based on Triangular Distributions Sangwon Eum, Korea University Business School, Seoul, Korea, Republic of, Hosun Rhim We propose a DEA-AR model using stochastic weight bounds. For practical purpose, triangular distribution is assumed. We test the model by changing parameters of the distribution. Real data set is used for the test. Chang Won Lee, Hanyang University, School of Business, 17 Haengdang Dong, Seoul, 133-791, Korea, Republic of, Jae Sun Jeong

210