2015 Informs Annual Meeting

WD18

INFORMS Philadelphia – 2015

3 - Pooling Principals by a Repair Agent Shuo Zeng, University of Arizona, McClelland Hall 430, 1130 E. Helen Street, Tucson, AZ, 85721, United States of America, shuozeng@email.arizona.edu, Moshe Dror The literature on principal-agent interplay has its focus on the principal. We focus on the agent. For performance based service contracts it is known that the principal extracts all the surplus and the agent breaks even. But this is not the case for an agent contracting with multiple principals. We show that agent who contracts with a collection of principals with interdependent failure characteristics realizes a profit rate that is convexly increasing in the number of principals. 4 - A Game Theoretic Study on Fake Goods Issue in C2c E-commerce This paper mainly researches the fake goods issue of C2C e-commerce sites by analysis on suppliers, buyers and sellers based on game theory. Supplier’s credit problem leads to the spread of fake goods, which effect seller’s credit and arouse fraud, caused by the imperfect credit management. Based on that, we build a game model,and then we make an improvement by regulating the suppliers. 5 - How Personality Type Changes the Impact of Recommendation on Investment Decisions Jun-Yuan Chen, Frontier High School, 1601 East Debbie Ln, Apt. 1301, Mansfield, United States of America, piky1223@gmail.com We study how personality types, measured by the Myer-Briggs Type Indicator (MBTI) survey, affect the impact of recommendations on investment choices. The MBTI measures personality types in multiple dimensions including how individual processes information, and react to stimulus. We manipulated the presence of a recommendation between two investment choices. We found the impact of this manipulation is mediated by the results of the personality type survey. Fen Ding, Huazhong University of Science and Technology, School of Management,1037 Luoyu Road, Wuhan, China, ivyours319@163.com Irina Cazan, Carnegie Mellon University, Electrical and Computer Engineering, 2134 Hamerschlag Hall, Pittsburgh, PA, United States of America, icazan@andrew.cmu.edu, Connor Walsh, Radu Marculescu This decade has seen broad research in the social behavior of microbes, with parallels drawn to human behavior to explain core network formation processes. This work explores static and dynamic properties of microbial and human social networks to identify the role of such processes in forecasting community evolution. Using the human microbiome and startup networks, similarities are identified in static features and reaction to disruptions, and contrasts in dynamic evolution. 2 - Graph Based Approaches for Managing Cyber-physical Systems Fabian Runge, Research Assistant, Jade University, Friedrich- Paffrath-Strasse 101, Wilhelmshaven, 26389, Germany, fabian.runge@jade-hs.de, Sabine Baumann, Oliver Eulenstein Common solutions for controlling cyber-physical systems focus on linear approaches. However, the system consists of self-referential, but connected and interacting systems and thus high complexity. Given the growing amount of related (sensor) data and the interdependencies of the systems, graph based solutions, like neural networks, seems to be more suitable for controlling the production. This paper describes current graph related methods to handle the growing complexity. 3 - On the Statistical Monitoring of Communication Networks Marcus Perry, University of Alabama, 305 Alston Hall, 361 Stadium Drive, Tuscaloosa, AL, 35487, United States of America, mperry@cba.ua.edu, Ketong Wang, Xuwen Zhu Often, decision-makers need to be aware of significant organizational changes in advance to avoid or mitigate potential crisis. Communication networks often serve as a proxy for assessing organizational structure. In this talk, we discuss application of statistical process control methods to efficiently detect changes in macro organizational structure within communication networks. We apply our approach to a time series of daily email networks from the Enron email corpus during crisis time. WD17 17-Franklin 7, Marriott Networks and Graphs I Contributed Session Chair: Boris Brimkov, Rice University, 6100 Main MS-134, Houston, TX, 77005, United States of America, bb19@rice.edu 1 - Micro to Macro Community Scaling in Human and Bacterial Societies

4 - Efficient Computation of Chromatic and Flow Polynomials Boris Brimkov, Rice University, 6100 Main MS-134, Houston, TX, 77005, United States of America, bb19@rice.edu, Illya Hicks The chromatic and flow polynomials of a graph count the number of ways to color and assign flow to the graph. We present closed formulas and polynomial- time algorithms for computing the chromatic polynomials of novel generalizations of trees, cliques, and cycles. We also use graph duality to compute the flow polynomials of outerplanar graphs and generalized wheel graphs.

WD18 18-Franklin 8, Marriott Optimization Robust II Contributed Session

Chair: Svenja Lagershausen, Leibniz Universitaat Hannover, Wirtschaftswissenschaftliche Fakultaat, Königsworther Platz 1, Hannover, 30167, Germany, svenja.lagershausen@prod.uni-hannover.de 1 - Robust Pessimistic Bi-level Optimization

Ihsan Yanikoglu, Özyegin University, Nisantepe Cekmeköy, Istanbul, Turkey, ihsan.yanikoglu@ozyegin.edu.tr, Daniel Kuhn This paper proposes a robust optimization approach for a class of pessimistic bilevel optimization problems with uncertain data. The associated optimization problem consists of binary ``here and now’’ decisions that are made before data reveals itself; continuous ``wait and see’’ decisions that are adjustable according to the revealed portion of the data. We propose conservative and progressive approximations of such bilevel optimization problems. 2 - Investor Avoidance from Risk as Uniform Portfolio Becomes Optimal Ahmed Burak Paç, PhD Candidate, Bilkent University, Department of Industrial Engineering, Ankara, 06800, Turkey, burakpac@gmail.com In a market of N risky assets, asset returns follow a multivariate distribution involving distributional uncertainty in a ball around a known nominal distribution. As the radius of uncertainty increases, optimal investment converges to the uniform portfolio with equal 1/N wealth on each asset. In this study, the tendency of the investor to respond to incresing uncertainty by avoiding risk, i.e., Jens Bengtsson, Associate Professor, School of Economics and Business, Norwegian University of Life Sciences, P.O Box 5003, Aas, 1432, Norway, jens.bengtsson@nmbu.no, Mikael Ronnqvist, Patrik Flisberg Several studies indicate relationships between changes in input prices and output prices in process industries, e.g. oil refinery. It is of interest to analyze how such relationships can be incorporated in uncertainty constaints which then is used in robust optimization of decisions in the supply chain. Then it also of interest to analyze how different uncertainty constraints will affect the planning of the supply chain, risk exposures and the cost of robustness. 4 - Robust Harvesting Planning in Lumber Supply Chains with Random Supply and Demand Omid Sanei Bajgiran, PhD Candidate, Concordia University, 1455 De Maisonneuve Blvd. W., Montreal, QC, Canada, o_sane@encs.concordia.ca, Mustapha Nourelfath, Masoumeh Kazemi Zanjani We propose a robust harvesting planning model under log supply and demand uncertainty that affect the right hand side, constraints, and the objective function coefficients. The proposed robust optimization model which has been formulated based on “price of robustness” provides some insights into the adjustment of the level of robustness of the harvesting plan over the planning horizon and protection against uncertainty. 5 - Dynamic Multi-product Lot-sizing Problem under Uncertainty Svenja Lagershausen, Leibniz Universität the uniform portfolio, is investigated, introducing a riskless asset. 3 - Robust Optimization of Process Industries under Price Uncertainty

Hannover, Wirtschaftswissenschaftliche Fakultät, Künigsworther Platz 1, Hannover, 30167, Germany, svenja.lagershausen@prod.uni-hannover.de

We present a stochastic single-level, multi-product dynamic lot-sizing problem subject to a strict production capacity constraint. The production schedule is determined such that the expected costs are minimized. The backlog is limited using a d-service-level constraint. This leads to a non-linear model that is approximated by a linearization model.

459