INFORMS 2021 Program Book

INFORMS Anaheim 2021

MC34

relationships where the micro-level team characteristics affect each other: while individuals and dyads affect the macro-level team performance, the rest of the team also affect the micro-level individual and dyad performance. Our study adds to the prior literature by showing the different effects of learning on individuals and dyads, which enhances the micro-level organizational design of hospitals. MC33 CC Room 209A In Person: Empirical Research in Retail Operations General Session Chair: Sahar Hemmati, University of Maryland, United States 1 - Strategic Visual Merchandising of New and Open-box Products: Evidence From Experiments and Retail Data Retailers are increasingly selling returned products as open-box along with their new counterparts, which raises the question of what’s the most effective visual merchandising strategy for this assortment? While some retailers position open- box products side-by-side with their new counterparts in the assortment (i.e., the side-by-side strategy), others position them separately in a different part of the retail space/different page on a website (i.e., the separate strategy). We conduct multimethodology research to empirically investigate the economic effectiveness of these two visual merchandising strategies. 2 - Impact of Price Markdown Framing on Product Returns Wedad Elmaghraby, University of Maryland, College Park, MD, United States, Sahar Hemmati, Ozge Sahin Percentage discounts and bundle discounts are among the mostly used marketing tools in retail, the impact of which on sales and customers’ purchase behavior has been extensively studied in the extant literature. However, the effect of different pricing strategies on product returns has not been well-explored. Our objective is to understand how percentage discounts and bundle discounts impact customers’ product choice and return decision. Using the data from one of the largest apparel retailers in Turkey, we find that bundle promotions not only increase the incidence, but also decrease the return probability of each product, controlling for price, discount depth and item characteristics. We find that returns of products purchased with a bundle discount decrease on average by 21% compared to returns of the same products while purchased with a percentage discount. MC34 CC Room 209B In Person: Service Science IBM Best Student Paper Competition (II) Award Session Chair: Yunzong Xu, Massachusetts Institute of Technology, Cambridge, MA, 02142, United States 1 - Ownership Utility Estimation in Rent-to-Own Businesses Milad Armaghan, University of Texas, Richardson, TX, 75080- 3021, United States Rent-to-own firms rent products in exchange for a periodic fee and offer the already-rented products for purchase at buyout prices to their renters. The renters’ ownership utility for a rented product determines their willingness to pay the buyout price for the product. To model renter decisions, we develop a utility framework that incorporates the unique features of the RTO business, namely the repeated signals about each renter’s utility from his responses to different buyout prices. Using transaction data from an RTO firm, we compare the estimation performance of utility specifications. 2 - Blind Network Revenue Management and Bandits with Knapsacks under Limited Switches Yunzong Xu, Massachusetts Institute of Technology, Cambridge, MA, 02142, United States We study both the classical price-based network revenue management problem in the distributionally-unknown setup, and the bandits with knapsacks problem. Beyond the classical resource constraints, we introduce an additional switching constraint to these problems, which restricts the total number of times that the decision-maker makes switches between actions to be within a fixed switching budget. For such problems, we show matching upper and lower bounds on the optimal regret, and propose computationally-efficient limited-switch algorithms that achieve the optimal regret. Yuanyuan(Amy) Ding, University of Minnesota, MN, United States, Necati Ertekin, Karen L. Donohue

MC31 CC Room 208A In Person: Advanced Machine Learning Techniques in Manufacturing Systems Joint Session Chair: Hao Yan, Tempe, AZ, 85281-3673, United States Co-Chair: Imtiaz Ahmed, Texas A & M University, College Station, TX, 77840-6717, United States 1 - Deep Multistage Multitask Learning for Quality Prediction Of Multistage Manufacturing Systems Hao Yan, Arizona State University, Tempe, AZ, 85281-3673, United States, Nurettin Dorukhan Sergin, William A. Brenneman, Shan Ba In multistage manufacturing system, modeling multiple quality indices based on the process sensing variables is important. However, the classic modeling technique predicts each quality variable one at a time, which fails to consider the correlation within or between stages. We propose a deep multistage multi-task learning framework to jointly predict all output sensing variables in a unified end- to-end learning framework according to the sequential system architecture in the MMS. Our numerical studies and real case study have shown that the new model has a superior performance compared to many benchmark methods as well as great interpretability through developed variable selection techniques. 2 - Surprise Driven Autonomous Experimentation Platform Imtiaz Ahmed, Texas A. & M, Plantation, College Station, TX, 77840-6717, United States Imtiaz Ahmed, West Virginia University, Morgantown, WV, United States, Yu Ding Physical experiments are often costly and refrain us from exploring high dimensional parameter spaces to find the most suitable design or approximating the underlying response surface . In this work, we develop an autonomous experimentation platform for discovering new, efficient design. We introduce the notion of ‘surprise’ and propose a surprise guided exploitation-exploration trade off policy. We combine our surprise driven active learning techniques with computer-controlled simulated experiments to guide the sequential physical experiment selection process. Our platform can plan and execute the sequential experiments autonomously and reach the desired design using minimal search and resource. We compare our approach with the Bayesian optimization based sequential experimentation policy to illustrate its benefits and potential applications. Chair: Abhishek Deshmane, IESE Business School, United States 1 - Reading Between the Stars: understanding the Effects of Online Customer Reviews on Product Demand Hallie Cho, Vanderbilt, Nashville, TN, 37203, United States, Manuel Sosa, Sameer Hasija Consumer perceptions of product quality—and how they are shared via customer reviews—are of extreme relevance to the firm, but we still do not understand how the quantitative and qualitative aspects of customer reviews affect product demand. Our paper seeks to fill this critical gap in the literature by analyzing star ratings, the sentiment of customer reviews, and their interaction. Using the US automobile market data, we find robust empirical evidence that 1) review sentiment and star ratings both have a decreasingly positive effect on product demand and 2) the effect (on demand) of their interaction suggests that the two components of reviews are complements. Positive sentiments in text reviews increase the positive effect of ratings when the effect of ratings is decidedly positive and they compensate for the tendency of consumers to discount extremely high star ratings. 2 - Exploring the Trilateral Productivity in Surgery Teams: Do Agents, Pairs, and Teams Affect Each Other? Jaeyoung Kim, PhD Candidate, Clemson University, Clemson, SC, United States, Lawrence Fredendall, Ahmet Colak, Robert Allen We study the effect of micro-level foundations of surgical teams on team productivity: interactions among individuals (agents), dyads (pairs), and teams. Using six datasets containing the micro-level characteristics of surgical teams (e.g., patient, operation, staff, procedure, scheduling, and surgeons) obtained from a large south-eastern hospital, we show that surgical teams have trilateral MC32 CC Room 208B In Person: Empirical Research in Operations Management General Session

67