Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

POSTERS

16 - Framing in Supplier Selection Problem Somaye Ramezanpour Nargesi, PhD Candidate, University of Texas at Arlington, Arlington, TX, 76015, United States, Kay-Yut Chen One of the crucial steps of procurement process is supplier selection. Assuming that the supply management of the company is single sourcing, the real problem will be a multi-attribute trade off decision to choose one supplier from several. In this research we tackled this problem from behavioral economy stand point. We administered experimental design to investigate the effect of two types of framing (Content and Information Framing) on this multi-attribute trade-off decision and found out that both types of framing matter in a meaningful way. 17 - Impacts of Government Subsidies on Cooperative Game Pricing for China Railway Express and Freight Forwarders Jingru Ren, Southwest Jiaotong University, Chengdu, China, Yinying Tang, Qinglin Li, Weidong Dai China’s subsidies to China-Europe express can be divided into subsidies for CHINA RAILWAY Express and freight forwarders.To study the impacts of subsidies on cooperative pricing strategy for the Express and forwarders,considering market demand fluctuation,the Stackelberg game model under different subsidy objects and Nash model under non-subsidy situation have been established.The optimal contract pricing strategy,freight volume and market pricing strategy have been obtained.A numerical example has been used to verify the validity of the models and to analyze the impacts of different subsidy policies on both sides’ revenue and supply chain revenue when market demand fluctuates. 18 - Project Team Communication and Understanding Cycles of Mental Model Convergence Deanna Kennedy, Associate Professor, University of Washington, Bothell, WA, 98011-8246, United States Project teams are thought to produce better quality outcomes than individuals working alone. Yet, given the inconsistency of teamwork effectiveness found by researchers for at least the last 35 years, there is still much to be learned about why teams achieve, or fail to achieve, more than the sum of their parts. Herein, I look at team mental models, cognitive repositories containing abstract representations relevant to taskwork and teamwork activities, and model the system to understand the dynamics of shared and accurate team mental models over time. This research uses agent based modelling and particle swarm optimization to study the system and provide preliminary analysis about teams over time. 19 - The Effects of Manager’s Perception on Intellectual Property, Innovation, and Firm Performance Giyoon Kwag, Illinois Institute of Technology, Chicago, IL, United States, Sejun Park, Taewoo Roh Analyzing survey data from CEOs collected by Science and Technology Policy Institute (STEPI) in South Korea, we establish further understandings on the Intellectual Property (IP) and its impacts in the perspective of CEOs. The research investigates how the CEOs’ attitude toward IP affects the firm performance through product, process, organization and marketing innovations. This study examines the direct and indirect effects of IP on innovation and financial performance with the structural equation method (SEM) and captures the mediating role of innovation between IP and financial benefits. 20 - Stochastic Programming-based Bypass Strategy for Acute Stroke Patients and EVT Resource Re-distribution Ting-Yu Liu, National Tsing Hua University, Hsinchu, Taiwan, Ming-Ju Hsieh, Chun-Han Wang, Yu-Ching Lee, Wen-Chu Chiang The prognosis performance of endovascular therapy (EVT) is shown better than thrombolysis in the patients with acute ischemic stroke (AIS) caused by large vessel occlusion in recent clinical medical research. Yet EVT resources are available only in a few number of comprehensive stroke centers. According to stroke symptom severity and geographic information, we develop a bypass strategy for AIS patients using a stochastic programming model to decide a comprehensive stroke center where the patient should be sent. Policy of reduction or expansion of EVT hospitals can further be facilitated. 21 - An Integrated Approach to Train Timetabling and Rolling Stock Circulation in Urban Rail Transit Lines Bum Hwan Park, Korea National University of Transportation, Uiwang, Korea, Republic of There have been many tries to reduce travel time in rail transit. One of them is to diversify the stop patterns to provide faster trains to the passengers. Diversifying the patterns makes it difficult to find an optimized plan since there are some operational constraints like overtaking, complicated headway to consider. And a rolling stock circulation can not be easily found due to different running times. We present a network model to consider timetabling and rolling stock circulation, simultaneously. We first find a circulation with available stocks, which may be not feasible. And then the model finds an optimal circulation with shifted arrival and departure time by column generation approach.

22 - Optimizing Green Infrastructure Placement Under Precipitation Uncertainty

Masoud Barah, University of Tennessee, Knoxville, TN, 37996- 2315, United States, Anahita Khojandi, Xueping Li, Jon Hathaway, Olufemi A. Omitaomu Green Infrastructure (GI) practices are low cost, low regret strategies that can contribute to urban runoff management. However, questions remain as to how to best distribute GI practices through urban watersheds given the precipitation uncertainty and the hydrological responses to them. We develop a stochastic programming model to optimize placement of GI practices in a watershed to minimize the excess runoff under medium-term precipitation uncertainties. The model was calibrated using precipitation projections and the associated watershed hydrologic response. The optimal GI placement is identified for an urban watershed in a mid-sized city in the U.S. 23 - A Three-level Defender-attacker-operator Problem Against Cyber-attacks in Electric-gas Systems Bining Zhao, Lehigh University, Bethlehem, PA, 18015, United States, Alberto Lamadrid, Rick Blum The interdependence between natural gas and power systems is increasing rapidly. Availability of natural gas for gas-fired units can impact the secure operation of power systems. Fuel supply shortage for gas-fired generators can be caused by uncertain interruptible supply and incorrect supply information. This work proposes a trilevel min-max-min optimization problem to provide the power system operator a practical tool to protect critical fuel supply information from man-in-the-middle cyber-attacks, and also the strategies to sign firm supply contract to reduce gas supply uncertainties. Column and constraint generation (C&CG) algorithm is employed to solve the proposed problem. 24 - Dandelion Algorithm: A New Meta-heuristic Based on the Wind Markov Chain Kihyuk Yoon, UNIST, Ulsan, Korea, Republic of, Jongkyung Shin, Chiehyeon Lim, Chiehyeon Lim This work proposes a new meta-heuristic algorithm called the dandelion algorithm (DA). The DA is inspired by the dispersal of dandelion seeds by wind. Such dispersal can be considered as a random walk for optimization and may be affected by flower height (i.e., objective value) as well. In the DA, the two essential phases of optimization, exploration and exploitation are achieved during the designed mechanisms of planting, growth and survival, and seed dispersal. Several engineering problems are used for evaluation, and the result is compared with those of several existing algorithms. Results show that the proposed DA is powerful. 25 - Algorithms for the Mean Steady-state Waiting Time in the GI/GI/1 Extremal Queue Yan Chen, Columbia University, New York, NY, 10027, United States, Ward Whitt The effective algorithms are developed to compute the steady-state mean waiting time of extremal GI/GI/1 queues. We establish two reductions for an extremal queue, reducing it to D/GI/1 and GI/D/1. The idle time simulation algorithms are exploited to esti- mate first two moments of idle time and then compute E[W] by associated equilibrium excess distribution of idle time. Also, the paper exploits fast numerical algorithms for the extremal queue, one is using the negative binomial recursive formula, and another is considering an equivalent discrete-time Markov chain recursion. The computational results for different cases of c2a,c2s are compared with known approximations and bounds. 26 - Dynamic Routing in a Many-server System with Costly Information Junfei Huang, The Chinese University of Hong Kong, Shatin, N.T., Hong Kong We study a multiple-station queueing system with a single class of customers and many servers in each several stations. Upon a customer’s arrival, the system manager decides which station the customer should be routed to. The station information retrieval is costly. We propose routing policies that are proved to be asymptotically optimal in minimizing the long-run average waiting cost plus the information retrieval cost. 27 - Revised Adaptive Linear Programming Algorithm Lin Guo, The University of Oklahoma, Norman, OK, 73071, United States We revise the Adaptive Linear Programming (ALP) algorithm. ALP is a linearization algorithm with limitations: its critical parameter Reduced Move Coefficient (RMC) is determined with heuristics and there is no mechanism of updating it to get more desired solution; little knowledge on the association between RMC and model behavior has been explored; no generic criteria for the designer to evaluate the quality of the results. To fill in the gaps, we revise the algorithm by using the insight obtained from post-solution analysis into the critical parameter updating to remove the heuristics. We use a test problem to examine the performance of the revised ALP algorithm and validate its advancement.

340