Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

POSTERS

41 - Addressing Parameter Ambiguity in Medical Decision Making Charmee Kamdar, University of Michigan, Ann Arbor, MI, United States Data driven mathematical models are increasingly being used for medical decision making. However, the parameters used in the mathematical models can be imprecise due to scholarly disagreement or small input data sizes. One way to account for this ambiguity in a Markov Decision Process (MDP) is to use a Multiple Model Markov Decision Process (MMDP) that has many different transition probability matrices. Using an MMDP results in a treatment policy that is more robust to changes in the transition probability matrix. This poster discusses how to construct MMDPs from observational data. 42 - A Clustering Based Decision Tree Approach for Risk Assessment of Information Security in Smart Cities Yana Yuan, PhD Candidate, Northwestern Polytechnical University, Xi’an, 710129, China, Huaqi Chai The emergence of new information technologies poses challenges to information security in smart cities. Risk is the most prominent challenge. In this paper, we analyze the impact factors of information security in smart cities and evaluate the risk that these factors bring to information security. We aim at establishing a new evaluation index system and building a risk assessment model based on Ward’s method and decision tree. Then we present an empirical study on risk assessment of information security in 15 smart cities. We provide a data-based and intelligent method to assess the risks of information security, and give solutions to address the challenges for information security in smart cities. 43 - Estimating and Analyzing Hazardous Material Flows in Oklahoma Ronny Pacheco, Oklahoma State University, Stillwater, OK, 74075, United States, Manjunath Kamath, Farzad Yousefian, Scott Frazier, Babak Farmanesh, Goutham Takasi We present a unique survey-driven hazardous material flow estimation model for the state of Oklahoma. Our approach begins with a database of Oklahoma commercial facilities, which store materials considered as extremely hazardous substances (EHS). A web-based survey targeting these facilities is used to collect shipment data for the EHS materials handled by these facilities. Shipments are then assigned to routes identified using a shortest path algorithm. A GIS application is being developed to analyze and visualize the flow patterns and intensities of the various EHS materials transported. The results will be useful to Local Emergency Planning Commissions in planning for emergencies. 44 - Efficiency Study in Public Tenders: Analysis of Framework Agreements in Chile. Eduardo Lara, Universidad de Chile, Santiago, 8370456, Chile, Marcelo Olivares, Gabriel Weintraub, Daniela Saban Framework Agreements (FA) is a commonly used procurement mechanism by governments and large organizations. It is based on an auction-type design to select an assortment of products from multiple suppliers with posted prices, allowing some flexibility and variety to purchasing units. The design of FA requires balancing competition to enter the market with the variety offered inside the market. This paper conducts an empirical study of the FAs used by the Chilean government, identifying inefficiencies in the procurement market, providing improvements to design and conducting a field study to measure the actual effectiveness of the new implemented design. 45 - The Effects of Video Game Players’ Emotions and Experiences on Online Review Ratings Pei-Hua Chen, National Chiao Tung University, 1001 University Road, Hsinchu, 300, Taiwan, Li-Chien Cheng Previous literature suggested that consumers prefer to make purchasing decisions based on online user-generated-contents. This study investigated the different effects of video game players’ emotions and game playing experiences on professional and general consumer reviews ratings for different kinds of video game genres. Text mining techniques was used to investigate the effects of polarity of online reviews, emotions and gaming experiences on video game sales on amazon.com. The results showed that factors affect review ratings for different genre of games are not the same. 46 - An Intelligent Design Assistant System Development: A Case of the Korean Shipbuilding Industry Jun-Mo Nam, Ulsan National Institute of Science and Technology This poster presents a development case of an intelligent design assistant system that can give an insight into; 1) optimal scheduling based on workload estimation; 2) response-bot based on past communication records with shipowners; and 3) design verification based on revision log analysis. We adopt various machine learning techniques including the gradient boosting and similarity matching algorithms to replicate experts’ decisions recorded in the legacy system. The system has been successfully integrated into the newly developed project management system and is expected to be deployed as a part of the Smart Shipyard Program of Korea by 2019. (UNIST), Ulsan, Korea, Republic of, Dong-Joon Lim, Yeo-Myung Rho, Je-Kyung Kim, Wook-Hyun Cha

47 - Modeling Human Decision Making Behavior in Simulation Farhad Moeeni, Professor, Arkansas State University, Jonesboro, AR, 72467, United States, Karen Yanowitz, John Seydel The purpose of this study is to acquire insights into the behavior of people when they encounter waiting lines. It is expected that several theories of decision- making to be examined by this study through controlled laboratory experiments. The first objective is to identify and possibly quantify the important factors that influences people’s behavior when they encounter waiting lines, for example balking and reneging. The second objective is, to apply what we learn, as a result of investigating the first objective, to optimize the queueing systems from the perspectives of both service receivers and service providers. 48 - Gaussian Mixture Model Based Random Search for Continuous Optimization via Simulation Wenjie Sun, Tongji University, Shanghai, China This paper studies integrated random search algorithms for continuous optimization-viasimulation (COvS) problems. We first tailor the Gaussian process- based search (GPS) algorithm to handle COvS problems. We then analyze the potential sampling issue of the GPS algorithm and propose to construct a desirable Gaussian mixture model (GMM) which is amenable for efficient sampling and at the same time also maintains the desirable property of exploitation and exploration trade-off. Then, we propose a Gaussian mixture model-based random search (GMRS) algorithm. We build global convergence of both the tailored GPS algorithm and the GMRS algorithm for COvS problems. 49 - An Integrated Simulation Optimization Framework For Controlling A Biological Invader An aggressive invasive plant, Sericea, reduces the abundance and diversity of native plants and causes large economic damage in the Great Plains of U.S. We present an integrated simulation-optimization framework to control this invader. Simulation mimics Sericea growth over space and 12 years, while a bio-economic optimization model finds an optimal search and treatment path to minimize its economic damage. 50 - Consolidation of Picked Orders in a Fulfillment Center with Explosive Storage Wen Zhu, New Jersey Institute of Technology, Newark, NJ, 07102, United States, Sanchoy Das Online fulfillment warehouses, similar to those used by Amazon, typically operate an explosive storage policy. That is, each item is stocked in multiple random locations dispersed throughout the warehouse. Orders are then picked and collected in totes which are assigned to one of many packaging stations. Each multi-item order is therefore located in several totes and must be consolidated or rebined at a packaging station. A perfect assignment is not possible. We formulate the problem a MIP and present a fast heuristic solution. 51 - Multi-attribute Supplier Selection Model Mazen Hussein, Assistant Professor, University of Wisconsin- Platteville, Platteville, WI, 53818, United States The main purpose of this work is to create a fuzzy logic model to select the most appropriate supplier based on the lowest reduction rate with multiple attributes. A mechanism with certain criteria is used to evaluate the suppliers and then select the best alternative among them. 52 - Multi-level Reverse Supply Chain Redesign Under Unknown Facility Capacity Jieyu Lei, Northwestern Polytechnical University, Xi’an, China Technological innovation motivates consumers, especially the young, to chase after new products constantly. Lots of old products has been put aside or dropped randomly. How to reuse these waste resources has been an urgent issue for many countries. This research aims to solve the problem by designing a multi-echelon reverse supply chain. We modeled a network optimization problem using the mixed integer planning approach and proposed a heuristic algorithm to calculate the optimal designing choice. The results show that through the model, we can not only get the optimal location and transportation scheme, but also the optimal capacity decision for new facilities. 53 - Estimate the Characteristics of Truck Flow Based on Spatial Scale Fadong Zhang, Southwest Jiaotong University, Chengdu, 610031, China, Mi Gan, Xinyuan Li This paper takes the truck trajectory data as the research object, classifies it by transportation efficiency index, and uses statistical physics method to probe the distribution of travel trajectory, transportation distance and freight volume of the individual truck, and the group trucks. At the same time, through the regression analysis of the relationship between the unload rate of trucks and the transportation distance and transportation capacity, the interaction characteristics of the probability distributions are explored, and the mechanism of the influence of truck mobility characteristics on the no-load rate of road freight is explained. Sevilay Onal, NJIT, Newark, NJ, 07103, United States, Najmaddin Akhundov, Esra B y ktahtak Jennifer Smith, Gregory R. Houseman

342