Informs Annual Meeting 2017

SB38

INFORMS Houston – 2017

SB39

3 - Service Design with Acclimation and Non-homogeneous Memory Decay Yifu Li, Hong Kong University of Science and Technology, Room 3208 , Lift 21, Clear Water Bay, Kowloon, 999077, Hong Kong, ylick@connect.ust.hk, Tinglong Dai, Xiangtong Qi In today’s “experience economy,” service providers increasingly emphasize creating memorable, delightful service experiences, a crucial driver of which is the schedule of activities in a service package. Empirical literature shows the most engaging activity should be scheduled neither at the beginning nor the end of the package. Theoretic literature, on the other hand, points to a U-shaped schedule. The paper closes this gap between empirical and theoretic literature and provides a foundation for advanced analytics that is called for in developing next- generation operating models.

352D Remembering Harvey Wagner Invited : INFORMS Special Sessions Invited Session

Chair Jayashankar M. Swaminathan, University of North Carolina, Kenan-Flagler Business School, Operations, Chapel Hill, NC, 27599- 3490, United States, msj@unc.edu, Co-Chair: Vinayak Deshpande, University of North Carolina, UNC, Chapel Hill, NC, United States, Vinayak_Deshpande@kenan- flagler.unc.edu 1- Remembering Harvey Wagner Jayashankar M. Swaminathan, University of North Carolina, Chapel Hill, NC, msj@unc.edu Harvey Wagner (1931-2017) had a profound impact on the field of Operations Research and Management Science during his lifetime. In this session, presenters will share their experiences with Harvey Wagner and how those influenced their thinking. 352E INFORMS Undergraduate Operations Research Prize Invited: Undergraduate Operations Research Prize Invited Session Chair: Murat Kurt, Merck & Co., Inc., 2306 Sherbrook Street, Apt: 2, Pittsburgh, PA, 15217, United States, murat.kurt7@gmail.com 1 - Fluid Optimal Control of Polling Systems with Large Switchover Times Yue Hu, Northwestern University, Evanston, IL, United States, yuehu2017@u.northwestern.edu, Ohad Perry, Jing Dong We study the optimal-control problem of stochastic polling systems with large switchover times. Under proper scaling, the stochastic system can be approximated by a Hybrid Dynamical System (HDS). For the HDS, we analyze two classes of controls, and prove that any (periodic) stationary behavior can be achieved by one of those controls. We employ this latter result to prove that the exhaustive policy is optimal among all stationary controls with respect to long- run average holding costs. 2 - A Generalized Markov Chain Model to Capture Dynamic Preferences and Choice Overload Agathe Soret, Columbia University, New York, NY, United States, acs2298@columbia.edu, Vineet Goyal Modeling preferences is a key problem in many settings. Many random utility based parametric models have been considered, but all suffer from two limitations: failure to capture i) dynamic preferences that depend on the offered assortment, and ii) the choice-overload phenomenon (where the probability of purchase decreases with the number of available options). We propose a generalization of the MNL model that addresses both these limitations. We show that our model can be estimated efficiently and also present a near-optimal algorithm for constrained assortment optimization. 3 - Optimization Strategies to Support Analyses and Decisions to Increase the Coverage of Cervical Cancer Screening Tests in Rural Areas of Cundinamarca, Colombia Maria Camila Patino Torres, Universidad de los Andes, Bogota, Colombia, mp.patino10@uniandes.edu.co, Ivan Mura, Camilo Gomez The barriers in the access to diagnosis and treatment of cervical cancer in Colombia are illustrative of the inequalities among population groups in the country. Through data analysis, the implementation of network flow and vehicle routing models, and economic evaluations, we assess the coverage of existing facilities in a large rural area, and propose alternatives to improve it. Our results show that cost-effective alternatives exist to provide access to screening for vulnerable populations. SB40

SB38

352C Recent Advances in Approximations for

Multi-objective Integer Programs Sponsored: Multiple Criteria Decision Making Sponsored Session Chair: Hadi Charkhgard, University of South Florida, USF, Tampa, FL, 33559, United States, hcharkhgard@usf.edu 1 - A Feasibility Pump and Local Search Based Heuristic for Biobjective Pure Integer Linear Programming Hadi Charkhgard, University of South Florida, Tampa, FL, 33559, United States, hcharkhgard@usf.edu, Aritra Pal We present a new heuristic algorithm to approximately generate the nondominated frontier of bi-objective pure integer linear programs. The proposed algorithm employs a customized version of several existing algorithms in the literature of both single-objective and bi-objective optimization. Our proposed method has two desirable characteristics: (1) There is no parameter to be tuned by users other than the time limit; (2) It can naturally exploit parallelism. A computational study shows the efficacy of the proposed method and also the value of parallelization on that. 2 - A Feasibility Pump and Local Search Based Heuristic for Solving Multiobjective Mixed Integer Linear Programs Aritra Pal, University of South Florida, 4202 E. Fowler Avenue, Tampa, FL, 33620, United States, aritra1@mail.usf.edu, Hadi Charkhgard We present a new heuristic algorithm to approximately generate the nondominated frontier of multiobjective mixed integer linear programs. The proposed algorithm employs a customized version of several existing algorithms in the literature of both single objective and multiobjective optimization. Our proposed method has two desirable characteristics: (1) There is no parameter to be tuned by users other than the time limit; (2) It can naturally exploit parallelism. An extensive computational study shows the efficacy of the proposed method on some existing standard test instances in which the true frontier is known. 3 - A New Algorithm to Optimize a Function Over the Set of Efficient Solutions for Biobjective Mixed Integer Linear Programming Alvaro Sierra, University of South Florida, 4411 Shady Terrace Lane, Unit A, Tampa, FL, 33613, United States, amsierra@mail.usf.edu, Hadi Charkhgard We present a new criterion space search algorithm for optimizing a function over the set of efficient solutions of a biobjective mixed integer linear program. The algorithm is easy to implement and because it maintains a lower and an upper bound on the value of the linear function at any point in time, it can be used to quickly generate a provably high-quality approximate solution. An extensive computational study shows that the efficacy of the algorithm. 4 - Goal Programming - A Comparison with Other MCDM Methods Nilakantan Sundara Raman Narasinganallur, K.J. Somaiya Institute of Management Studies & Research, B.602 Tulip Rachna Garden, Mulund Colony, Mumbai, 400082, India, nilakantan@somaiya.edu For solving multi-criteria decision problems, different schools of thought have developed around different methodologies. Goal Programming is an elegant multi-objective programming technique using the structure of mathematical programming and extending it beyond its confines of single objective situations. Due to its simplicity and structure, it is much better than other multi-objective, multi-criteria decision-making techniques like Analytical Hierarchy Process, Fuzzy Theory etc. The presentation will discuss a few examples of Goal Programming and evaluate its merits vis-à-vis other MCDM techniques.

48

Made with FlippingBook flipbook maker