2015 Informs Annual Meeting

WE29

INFORMS Philadelphia – 2015

WE26 26-Room 403, Marriott Project Management II Contributed Session Chair: Fang Xie, PhD Student, Beihang University, 37 Xueyuan Road, Haidian District, Beijing, 100191, China, xiefangmm@163.com 1 - Estimation of Resource Allocation Patterns in a Portfolio of Engineering Projects Vishwanath Hegde, California State University East Bay, 25800 Carlos Bee Blvd, Hayward, CA, 94542, United States of America, vish.hegde@csueastbay.edu, Zinovy Radovilsky Using historic resource loading data in a multi-project setting, we show that resource distribution patterns can be captured by parametric regression models, which can forecast resource distribution during project lifetime using project due date and other attributes. 2 - Improved Design of CMS by Considering Operators Primary and Backup Decision-making Styles Mohammad Rezaei-Malek, University of Tehran, No. 3, Ganji Alley, North Khosh Street, Tehran, 1457813353, Iran, m.rezaeimalek@ut.ac.ir, Reza Tavakkoli-Moghaddam, Nima Salehi Sadghiani This paper considers decision-making style (as an index of operator’s personal characteristics) in CFP to design an operator-consistent CMS. Decision-making style not only influences the interaction of two operators, but also affects the work that operator does on a machine, and these interactions both need to observe consistency. Hence, this paper presents a mathematical model for CFP that considers consistency between each two operators and consistency between operator and assigned task. 3 - Reactive Project Scheduling with a Cash Flow Balanced Objective Minjing Ning, Xi’an Jiaotong University, No.28, Xianning West Road, Xi’an, China, ningminjing@stu.xjtu.edu.cn, Zhengwen He This paper investigates reactive project scheduling which may be used to repair project schedules that suffer from multiple activity duration disruptions during project execution. The objective is to minimize the cumulative cash flow gap of the contractor in the real executing process of the project. 4 - Robust Scheduling of the Resource-constrained DTCTP with Uncertain Activity Costs Fang Xie, PhD Student, Beihang University, 37 Xueyuan Road, Haidian District, Beijing, 100191, China, xiefangmm@163.com, Zhe Xu We investigate the resource-constrained discrete time/cost trade-off problem in which the activity costs are stochastic and the objective is to construct a robust baseline schedule that maximizes the probability of completing the project within the given budget. Two algorithms for solving this problem are presented. We compare the two algorithms and analyze the impact of different factors through conducting experiments on a set of instances generated from the PSPLIB. 5 - Project Management And Quality Data Challenges for IoT Systems Michael Chuang, SUNY - New Paltz, 1 Hawk Dr, New Paltz, United States of America, chuangm@newpaltz.edu, Kuan-Tsae Huang Internet of Things (IoT) has shown its potentials to be employed to scenarios of Industry 4.0. Caterpillar installs sensors and telematics in its products. AzTrong uses embedded sensors to allow for bidirectional communication over production lines. Collected data enable remote repair and service to make appropriate deci- sions, resulting in increased manufacturing uptime and improved customer serv- ice. How to apply project management to manage IoT becomes an important but uncharted topic.

We develop a two-stage decision making model on how a governmental agency can build and incentivize its energy policy across service providers. The agency, as the Stackelberg leader, has multiple objectives related to economic concerns, environment, and energy surety. The agency sets penalties and limits to which the power companies respond by choosing their investment and production strategies. We develop a model that integrates the Successive Weighted Sum method into the policy making game. 2 - A Heuristic Based on Qualitative Information for Territorial Partitioning Problems Salem Chakhar, Dr, Portsmouth Business School, Portland Building, Portland Street, Portsmouth, PO1 3AH, United Kingdom, salem.chakhar@port.ac.uk, Maria Barbati, Carmela Piccolo, Giuseppe Bruno This presentation proposes a heuristic to solve territorial partitioning problems. It uses as input a tree data structure, previously constructed based on qualitative information. This qualitative evaluation is grounded on several criteria and takes the form of a qualitative scale with a finite set of evaluation levels. The heuristic is illustrated using real-world data relative to Ile-de-France region in France. 3 - A Bi-Level Decentralized Programming for Setting Differential Subsidy Rate of Taiwan’s Waste Printer This study uses bi-level decentralized programming for setting differential subsidies on the recycling plants in Taiwan. The case of waste printers is illustrated. The results show that the differential subsides on recycling plants can achieve a higher recycling rate. 4 - Markov Method for Assessing Utility Functions Baback Vaziri, Purdue University, 315 N. Grant St., West Lafayette, IN, United States of America, bvaziri@purdue.edu, Yuehwern Yih, Tom Morin, Mark Lehto Multiattribute value functions are a useful tool for decision makers. Many methods directly obtain information from the decision maker regarding the preferences of attributes. We propose an alternative approach, which will reverse engineer the weights of the value function. We use the results of the preferences of alternatives in conjunction with a Markov-based ranking method to develop a rating vector of attributes. 5 - Operations Scheduling in Reverse Supply Chains: Delivery Deadlines and Identical Demand Gang Wang, Assistant Professor, UMass Dartmouth, 285 Westport Road, Room 214, CCB, North Dartmouth, MA, 02747, United States of America, gwang1@umassd.edu, Angappa Gunasekaran This study addresses an integrated operations scheduling problem of reverse supply chains with delivery deadlines. The problem is to determine shipping quantities from collectors to the manufacturer and the assignment of demand points, subject to the capacity constrains of both the collectors and the manufacturer. Jiun-Yu Yang, Master Student, Tamkang University, 151 Yingzhuan Rd., New Taipei, 25137, Taiwan - ROC, jiunyu.yang@gmail.com, Hsu-shih Shih

WE29 29-Room 406, Marriott Big Data: Inference and Prediction

Sponsor: Analytics Sponsored Session

Chair: Rob Lantz, Senior Manager Of Operations Analysis, Novetta Solutions, 8618 Westwood Center Drive, Suite 315, Vienna, VA, 22182, United States of America, rlantz@novetta.com 1 - Detecting Unknown Threats through Social Network Analysis Matt Teschke, Senior Quantitative Consultant, Novetta, 7921 Jones Branch Drive, 5th Floor, McLean, VA, 22102, United States of America, mteschke@novetta.com, Jennifer Stave A common impediment to the analysis of networks is the determination of risk relative to particular actors within the network. Using insights from the field of SNA in addition to an understanding of the challenges faced by the analyst, entities can be prioritized for investigation. This network-centric approach assigns risk based on an assessment of an entity’s characteristics and activity using an eigenvector centrality algorithm, of which Google’s PageRank algorithm is one application.

WE27 27-Room 404, Marriott Multicriteria Decision Making II Contributed Session

Chair: Gang Wang, Assistant Professor, UMass Dartmouth, 285 Westport Road, Room 214, CCB, North Dartmouth, MA, 02747, United States of America, gwang1@umassd.edu 1 - A Game Theoretic Approach to Energy Policy Making with Multiple Objectives Busra Keles, University of Miami, 1251 Memorial Drive,

Department of Industrial Engineering, Coral Gables, FL, 33146, United States of America, bxk96@miami.edu, Murat Erkoc, Nurcin Celik, Mahide Kucuk, Yalcin Kucuk

487