2015 Informs Annual Meeting

MB24

INFORMS Philadelphia – 2015

MB24 24-Room 401, Marriott Data Mining and Network Inference for Social and Health Application II Sponsor: Artificial Intelligence Sponsored Session Chair: Sung Won Han, New York University, 650 First Avenue, New York, NY, United States of America, sungwonhan2@gmail.com Co-Chair: Chen Kan, University of South Florida, 4202 E. Fowler Ave. ENB118, Tampa, FL. United States of America, chenkan@mail.usf.edu 1 - Optimizing Display Advertising in Online Social Networks Zeinab Abbassi, PhD Candidate, Columbia University, 1214 Amsterdam Ave, 450 CSB, New York, NY, 10027, United States of America, za2153@columbia.edu Conventional online advertising methods need to be customized for OSNs. We propose probabilistic models and study the problem: given a number of impressions, what is the optimal order of users to show the ad to, to maximize the expected number of clicks? We show that this problem is hard to approximate. Therefore, we propose several heuristic algorithms. We evaluate the performance of these heuristics on real data sets, and observe that our two-stage heuristic outperforms baselines. 2 - Blood Donation Tailoring Problem to Improve Blood Supply Management Guven Kaya, PhD Student, Industrial Engineering, University of Blood donation tailoring is to identify blood donation types and collect blood products. Donors perform donation types that provide blood products to patients, having collection/inventory/spoilage costs. We collect data about donation types with demand, cost, time, eligibility percentages, compatibility from blood banks. We develop MIP models to find collected/spoiled/carried blood product amount on single and multi-period settings. We provide results based on data from blood donation centers. MB26 26-Room 403, Marriott Data Analytics Applications for Smart Industries Cluster: Globalization and International Activities Invited Session Chair: Grace Lin, Data Analytic Technology and Applications (DATA), Data Analytic Technology and Applications (DATA), Taipei, Taiwan - ROC, gracelin@iii.org.tw 1 - Is the Conventional Association Analysis Practical for Big Data Analytics? New Perspectives on Application and Computation Association analysis has been proven an NP-Complete problem. Owing to the inevitable challenge, it is necessary to devise an alternative way of discovering representative patterns. We present relative patterns discovery (named RPD) for big data analytics, which possesses four features: effectiveness, efficiency, panorama, and scalability. 2 - Towards Industry 4.0: Applying Big Data Analytics to Improve Manufacturing Performance Fish Yu, Data Analytics Technologies & Applications Research Institute, Institute for Information Industry, Taiwan - ROC, fishyu@iii.org.tw As a step towards the development of cyber-physical systems which play an important role in the transformation of manufacturing industry to the next generation known as Industry 4.0, this talk describes a log analytics framework that is capable of collecting, managing and analyzing large amount of machine data to enable real-time and predictive decision-making across various manufacturing processes. Experimental results using realistic data from semiconductor packaging tools show the effectiveness of the proposed framework. 3 - Green Multi-temperature Logistics using Time-dependent Data Analysis Wei-Ting Chen, Data Analytics Technologies & Applications Research Institute, Institute for Information Industry, Taiwan - ROC, weitingchen@iii.org.tw Multi-temperature food logistics contributes a considerable amount of greenhouse gas due to fuel burn and HFCs and PFCs generated by refrigeration. In this talk, Houston, E206 Engineering Bldg 2, Houston, TX, 77204, United States of America, gkaya@central.uh.edu, Ali Ekici Hao-Ting Pai, Data Analytics Technologies & Applications Research Institute, Institute for Information Industry, Taiwan - ROC, htpai@iii.org.tw

we will introduce how to estimate emissions depend on various levels of traffic condition, temporal demand patterns, delivery time windows, and different temperature control techniques. It helps carriers to respond to green policies of governments. 4 - Emerging Trends in ICT: using Big Data Analytics to Infuse New Energy into Smart Tourism Industry Tim Lin, Data Analytics Technologies & Applications Research Institute, Institute for Information Industry, Taiwan - ROC, timlin@iii.org.tw As many leading global organizations have applied Big Data Analytics to various public and commercial areas, valuable applications such as consumer insight, business operations optimization, and service innovation have been continuously increasing. In this talk, we will introduce a smart tourism solution which provides tourists real-time, personal, and proactive services by leveraging Big Data Analytics, resulting in a deep and authentic experience. The developed solution can support tourism-related businesses to connect with prospective customers and build responsive, efficient, and health smart cities and homelands.

MB27 27-Room 404, Marriott Advances in Multiobjective Programming Sponsor: Multiple Criteria Decision Making Sponsored Session

Chair: Margaret Wiecek, Department of Mathematical Sciences, Clemson University, Clemson, SC, 29634, United States of America, wmalgor@clemson.edu 1 - Parametric Simplex Algorithm for Linear Vector Optimization Problems Firdevs Ulus, Princeton University, ORFE, Sherrerd Hall, Princeton, NJ, 08544, United States of America, fulus@princeton.edu, Birgit Rudloff, Robert Vanderbei A parametric simplex algorithm for linear vector optimization problems is proposed. The efficiency of the algorithm is compared with Benson’s algorithm and the multiobjective simplex algorithm. For nondegenerate problems it outperforms Benson’s algorithm and is on par with the multiobjective simplex algorithm. For degenerate problems Benson’s algorithm excels the simplex-type algorithms; however, the proposed algorithm performs much better than the multiobjective simplex algorithm. 2 - An LP-based Branch-and-bound Algorithm for Biobjective Mixed Integer Programs Nathan Adelgren, Clemson University, Department of Mathematical Sciences, Clemson, SC, 29634, United States of America, nadelgr@g.clemson.edu, Akshay Gupte We introduce a new LP-based branch-and-bound (BB) method for solving biobjective mixed integer linear programs (BOMILP). New branching, fathoming, cutting plane and node relaxation techniques are incorporated into a traditional BB framework. Computational results show that this method is competitive with current techniques for BOMILP. 3 - The Quadrant Shrinking Method for Solving Triobjective Integer Programs Martin Savelsbergh, Prof, H. Milton Stewart School of Industrial & Systems Engineering, Georgia Institute of Technology, 765 Ferst Dr NW, Atlanta, GA, 30332-0205, United States of America, martin.savelsbergh@isye.gatech.edu, Hadi Charkhgard, Natashia Boland We present a new variant of the full (p-1)-split algorithm for finding all nondominated points of a triobjective integer program. The algorithm is easy to implement and solves at most 3n+1 integer programs, where n is the number of nondominated points. Computational experiments demonstrate its efficacy. 4 - Optimizing a Linear Function Over the Efficient Set of a Multi-objective Integer Program Hadi Charkhgard, University of Newcastle, University Drive, Callaghan, Australia, hadi.charkhgard@gmail.com, Natashia Boland, Martin Savelsbergh We present a new algorithm to optimize a linear function over the set of efficient solutions of a multi-objective integer program. Because the algorithm maintains both a lower and an upper bound on the optimal objective value, it can easily be converted into a fast approximation algorithm. Finally, we demonstrate that the algorithm can be used to efficiently compute the nadir point of a multi-objective integer program.

180