2015 Informs Annual Meeting

WC17

INFORMS Philadelphia – 2015

WC17 17-Franklin 7, Marriott Optimization Network Contributed Session Chair: Charles Nicholson, Assistant Professor, University of Oklahoma, 202 W. Boyd St., Room 116-H, Norman, OK, 73071, United States of America, cnicholson@ou.edu 1 - 1 Refueling Location Problem on a Continuous Tree Network Sang Jin Kweon, Ph. D. Student, The Pennsylvania State University, 232 Leonhard Building, The Pennsylvania State University, University Park, PA, 16802, United States of America, svk5333@psu.edu, Jose A. Ventura, Seong Wook Hwang In this talk, we consider a location problem for 1-refueling station on a tree network. We aim to locate the station anywhere on the network, including along the edge. Our objective is to locate a 1-refueling station to maximize the total traffic flow covered in round trips. For this, we derive reduction properties regarding the problem size and optimality conditions. Then, we develop an exact algorithm to determine the set of optimal locations for the refueling station. 2 - Near Optimal Design of Wavelength Routed Optical Networks Prahalad Venkateshan, Associate Professor, Indian Institute of Management, Ahmedabad, Vastrapur, Ahmedabad, India, prahalad@iimahd.ernet.in, Yogesh Agarwal The problem of designing a wavelength routed optical transport network is considered. Valid inequalities are used to augment traditional network design formulations. The resulting network is optimal for a majority of the problem instances tested and in those cases where it is not, a trial-and-error method finds near-optimal solutions. Computational tests are reported on relatively larger problem sizes than have been reported in literature on the wavelength routing problem. 3 - Metrics of Compactness: The Benefits of a Cluster-first, Route-second Approach to the Min-max K WRPP Oliver Lum, University of Maryland, 3103 Mathematics Building, University of Maryland, College Park, MD, 20742, United States of America, oliver@math.umd.edu, Bruce Golden, Edward Wasil, Carmine Cerrone In practice, it is often desirable for the routes of vehicles to exhibit certain properties that are not included in the objective function. Two such properties are compactness and separation. We discuss existing metrics that attempt to capture these traits, propose a new metric that combines these intuitions into a single value, and compare computational results for route-first, and cluster-first heuristics for the Min-Max K Windy Rural Postman. 4 - Optimal Flow Analysis, Prediction and Application Charles Nicholson, Assistant Professor, University of Oklahoma, 202 W. Boyd St., Room 116-H, Norman, OK, 73071, United States of America, cnicholson@ou.edu, Weili Zhang The fixed charge network flow problem is a classic NP-hard problem with many applications. To the best of our knowledge, this is the first paper that employs statistical learning technique to analyze the characteristics of optimal solutions. We develop an accurate propensity model based on this analysis to predict which arcs will have positive flow in an optimal solution. This propensity score allows for multiple applications such as identification of critical components in complex networks.

2 - A Robust Optimization Inventory Model with Uncertain Demand and Lead Time Mohammad Rahdar, Iowa State University, 133 University Village, Unit F, Ames, IA, 50010, United States of America, rahdar@iastate.edu, Lizhi Wang, Guiping Hu This study aims to improve the efficiency of a supply chain system by optimizing the ordering strategy in a manufacturing facility with uncertain demand and lead- time. A robust optimization model which has tri-level structure is proposed to explicitly address the uncertainty, and make planning decisions which are robust against all scenarios. We propose an exact algorithm for the tri-level programming model and report the numerical results. 3 - An Optimal Solution to a PCA Based Multi-Response Problem Nasser Fard, Associate Professor, Northeastern University, 334 Snell Engineering Center, 360 Huntington Avenue, Boston, MA, 02115, United States of America, n.fard@neu.edu, Huyang Xu, Yuanchen Fang It is shown that each eigenvalue obtained from the application of PCA in multi- response optimization has a set of eigenvectors, and different eigenvectors corresponding to each eigenvalue will lead to different results. A method for determining the optimal eigenvector combination is proposed. These eigenvectors are used to calculate the multi-response performance indices, and then the Zhi CHEN, Department of Decision Sciences, PhD B1-02 Biz2 Building, NUS Business School, Singapore, 117592, Singapore, chenzhi@u.nus.edu, Melvyn Sim, Xu Huan We investigate semi-infinite ambiguity sets that involve infinite expectation constraints besides moments and support. We study the associated intractable distributionally robust optimization by considering relaxed ambiguity sets with finite constraints. Based on worst-case distribution, we demonstrate an algorithm that tightens the relaxation gradually. We present expressive examples of these ambiguity sets, and show examples where the worst-case distribution can be relatively easy verified. 5 - Robust Optimization for a Location Problem with Codependent Uncertain Constraints Juan Carlos Espinoza Garcia, ESSEC Business School, Av Bernard Hirsch, CS 50105, Cergy-Pontoise, 95021, France, juancarlos.espinoza@essec.edu, Laurent Alfandari We consider an uncertain problem where the same parameter is liable to variation in both constraints and objective function. We model this problem under the Gamma-Robustness paradigm adapting the framework to include codependency between the uncertain equations. We apply this model to a location problem with uncertain demands and conduct computational experiments on a housing problem. WC19 19-Franklin 9, Marriott Heuristics Sponsor: Computing Society Sponsored Session Chair: Michael Hahsler, SMU, P.O. Box 750123, Dallas, TX, 75275, United States of America, mhahsler@lyle.smu.edu 1 - Heuristic Search of Good Decisions for Evaluating Soft Constraints in Integer Programs Steve Kimbrough, University of Pennsylvania, 3730 Walnut Street, Philadelphia, PA, 19104, United States of America, kimbrough@wharton.upenn.edu, Monique Guignard-Spielberg, Frederic Murphy Parameter sweeping seeks insight into (usually simulation) models by examining a plurality of model results obtained using a number of judiciously chosen parameter settings. We introduce decision sweeping (of optimization models), in which we collect a number of judiciously chosen decisions (feasible and infeasible settings of the decision variables) from the larger space of decisions. We focus on the resulting insights, especially with regard to integer programs with soft constraints. 2 - Ordering Objects: What Heuristic Should We Use? Michael Hahsler, SMU, P. O. Box 750123, Dallas, TX, 75275, United States of America, mhahsler@lyle.smu.edu Seriation, i.e., finding a suitable linear order for a set of objects given data and a merit function, is a basic combinatorial optimization problem with applications in modern data analysis. Due to the combinatorial nature of the problem, most practical problems require heuristics. We have implemented over 20 different heuristics and more than 10 merit functions. We will discuss the different methods in this presentation and compare them empirically using datasets from several problem domains. optimal factor-level combination will be uniquely determined. 4 - Distributionally Robust Optimization with Semi-infinite Ambiguity Sets

WC18 18-Franklin 8, Marriott Optimization Robust I Contributed Session

Chair: Zhi Chen, Department of Decision Sciences, PhD B1-02 Biz2 Building, NUS Business School, Singapore, 117592, Singapore, chenzhi@u.nus.edu 1 - Competitive Difference Analysis of One-way Trading with Limited Information Yingjie Lan, Peking University, Guanghua School of Management, Beijing, 100871, China, ylan@pku.edu.cn, Wei Wang We consider robust one-way trading with limited information on price as a game. Our analysis finds the best guarantee of difference from the optimal performance. We provide closed-form solution, and reveal for the first time all possible worst case scenarios. Numerical experiments show that our policy is more tolerant of information inaccuracy than Bayesian policies, and can earn higher average revenue than other robust policies while keeping a lower standard deviation.

432

Made with