Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

SB67

n SB67 West Bldg 105B

n SB68 West Bldg 105C

Joint Session ISS/Practice Curated: Digital Innovation, Analytics, and Business Value Sponsored: Information Systems Sponsored Session Chair: Jessica Pye, Arizona State University 1 - Using Swarm Optimization Heuristics for Winner Determination in Combinatorial Auctions Combinatorial auctions are increasingly being used to allocate bundles of items among interested bidders. However, as the number of bundles increases, it takes exponentially longer to solve for the winners in the auction. In such situations, regular solvers such as IBM CPLEX or AMPL fail to produce the optimal solution either completely due to computational limitations or in reasonable time. We propose an Ant Colony-based algorithm (TrACA) that produces optimal or near- optimal results within specified time. The best value from TrACA is at least as good as best-in-class heuristics in 100 % of the cases. Additionally, to better understand the why deterministic algorithms fail or take too long for approximating winners in such auctions, we analyze empirical hardness of tested instances and use a standard supervised learning method to build a predictive model of TrACA run-times. 2 - On the Networks of Executives and Firm Strategic Entry: Evidence from the US Broadband Industry Tedi Skiti, Temple University, Philadelphia, PA, 19123, United States, Paul Pavlou We examine the role of the social and professional networks of the firm’s top executives in the firm’s geographic entry to new markets. We define networks as social ties (generated by common education and societal activities) and professional ties (generated by common work experiences). Potential entrant firms may have an information asymmetry about the market conditions which might reduce their entry likelihood. Networks may reduce this information asymmetry providing information about competitors and market demand. We combine and utilize firm-level data on broadband deployment in the United States (US) and board member data about their networks and individual characteristics. We find that professional ties mitigate the competitive effects, making entry significantly more likely. We contribute to the limited literature on the role of the network of executives in a firm’s strategic entry (before the firm becomes an actual entrant) and the literature on intraorganizational knowledge transfer as the result of individual executives’ ties. 3 - When Your Problem Becomes My Problem: The Impact of Airline IT Disruptions on On-Time Performance of Competing Airlines Brad N. Greenwood, University of Minnesota, 4200 32nd St S, Arlington, VA, 22206, United States, Jennifer Tae, Min-Seok Pang We study how firm disruptions affect competitor performance in the presence of shared resources. We propose that the impact of disruption is moderated by the routine complexity of both a disrupted firm and a competitor. Our context is the four large IT outages affecting the U.S. airline industry. Competitor flights which originated from, or were inbound to, a disrupted hub experienced significant changes in on-time performance, depending on the routine complexity of the disrupted airline. Performance deteriorated during the disruption of full-service carriers, but improved during that of a low-cost carrier. We also find that this effect is strongly moderated by competitors’ routine complexity. 4 - Chat More and Contribute Better: An Empirical Study of a Knowledge-based Community Christopher M. Forman, Cornell University, Dyson School of Economics, Warren Hall, Ithaca, NY, 14850, United States, Xiaomeng Chen, Michael Kummer Platform design is important to the success of knowledge-based communities. This paper studies whether a new channel of information exchange can impact the efficiency of users’ contributions. To test this hypothesis, we use a policy change implemented on the platform. On 2010 October 15, Stack Overflow launched chat.stackoverflow.com to support real-time communication for users. We leverage the fact that not all users could initiate this communication channel to set up a difference-in-differences (DiD) strategy, that compares the questions of the same users before and after the exogeneous change. We find mixed evidence of the efficacy of the chat rooms on improving knowledge-sharing outcomes. Karthik Kannan, Purdue University, 3315 Webster Street, West Lafayette, IN, 47906, United States, Abhishek Ray

QSR Best Refereed Paper Award Sponsored: Quality, Statistics and Reliability Sponsored Session Chair: Haitao Liao, Ph.D., University of Arkansas, University of Arkansas, Fayetteville, AR, 72703, United States Co-Chair: Tirthankar Dasgupta, Ph.D., Rutgers University, 110 Frelinghuysen Rd, Piscataway, NJ, 08854, United States 1 - Change-Point Detection by Multivariate Adaptive Regression Weihong Guo, Rutgers, The State University of New Jersey, Piscataway, NJ, United States 2 - Minimizing Negative Transfer of Knowledge in Multivariate Gaussian Processes: A Scalable and Regularized Approach Raed Al Kontar, Garvesh Raskutti, Shiyu Ahou, University of Wisconsin, Madison, WI, United States 3 - Optimal Placement of Actuators for Composite Fuselage Shape Control Jianjun Shi,Georgia Institute of Technology, Atlanta, GA, United States, Juan Du, Xiaowei Yue, Jeffrey H. Hunt, 4 - Efficient Gaussian Process Prediction using Design-Based Subsampling Ying Hung, Rutgers State University of New Jersey, Piscataway, NJ, United States, Linglin He Joint Session QSR/Practice Curated: Data Science and Analytics for Quality and Reliability Assurance Sponsored: Quality, Statistics and Reliability Sponsored Session Chair: Mingyang Li, Tampa, FL, 33647, United States Co-Chair: Jianguo Wu, Peking University, Haidian District, Beijing 1 - Prescribing Age Replacement Policies with Minimal Repair in Anticipation of Non-stationary Unpunctual Implementation Shadi Sanoubar, University of Pittsburgh, Pittsburgh, PA, United States, Kai He, Lisa M. Maillart, Oleg A. Prokopyev We consider the problem faced by a maintenance planner who must prescribe an age-replacement policy in anticipation of the fact that the party implementing the policy may be unpunctual, to a degree that depends on the replacement age itself. Our analytical and numerical results compare the optimal solution and its performance to those when the unpunctual behavior is assumed to be absent or independent of the prescribed replacement time. 2 - Optimal Maintenance Management of the Deteriorating Water and Transportation Infrastructures with Interdependencies Hung Nguyen, University of South Florida, Tampa, FL, 33613, United States, Noha Abdel-Mottaleb, Shihab Uddin, Qiong Zhang, Qing Lu, He Zhang, Mingyang Li Due to the co-location and spatial proximity, the deteriorating water infrastructure (WI) and transportation infrastructure (TI) are physically and operationally interdependent/dependent. Most of the existing maintenance works mainly consider WI and TI separately, but neglect the complex interdependency/dependency between WI and TI. We will propose an optimal maintenance decision-making framework by jointly prioritizing and maintaining a large number of co-located WI and TI components (e.g., pipes and roads) at the reduced cost. A case study will demonstrate the cost-effectiveness of the proposed work. 3 - Degradation Modeling and Rul Prediction Using Wiener Process Subject to Multiple Change Points and Unit Heterogeneity Yuxin Wen, University of Texas-El Paso, 216 W. California Ave, El Paso, TX, 79902, United States, Jianguo Wu, Devashish Das, Bill Tseng Degradation modeling is critical for health condition monitoring and remaining useful life prediction (RUL). In the paper, we propose a multiple change-point Wiener process as a degradation model. To take into account the between-unit heterogeneity, a fully Bayesian approach is developed where all model parameters are assumed random. At the offline stage, an empirical two-stage process is proposed for model estimation, and a cross-validation approach is adopted for model selection. At the online stage, an exact recursive model updating algorithm is developed for online individual model estimation, and an effective Monte Carlo simulation approach is proposed for RUL prediction. n SB69 West Bldg 106A

56