2016 INFORMS Annual Meeting Program

POSTER SESSION

INFORMS Nashville – 2016

Attraction Recommendation For Tour Operators Under Stochastic Demand Chao Huang, Southeast University, Nanjing, Jiangsu Province, China, huangchao@seu.edu.cn, Weihao Hu The main stakeholders of attraction recommendations include tourists who refer to the recommendation to make traveling decisions and the tour operator who operates the recommendation. Existing attraction recommendation methods focus on recommending most relevant attractions to the tourists, yet overlook the benefit attraction recommendation could bring to the tour operator under stochastic demand. We provide a focused study on cost-based attraction under stochastic demand from the perspective of tour operators. We analyze the costs of attraction recommendation of tour operators and the impact of stochastic tourist demand, and formulate the cost-based attraction recommendation problem. We then propose a two-stage stochastic optimization model that involves joint chance constraint to optimize the attraction recommendation solution that focuses solely on tourist preferences, and solve the optimization model with Sample Average Approximation method. To verify the effectiveness of the proposed cost-based attraction recommendation method, comprehensive experimental studies are conducted with simulated instances as well as real-world case. Dynamic Pricing Of A Bottleneck With Heterogeneous User Preferences And Value Of Time Using Tradable Credits Mahyar Amirgholy, Postdoctoral Research Associate, Cornell University, 220 Hollister dr, Ithaca, NY, 14850, United States, amirgholy@cornell.edu, H. Oliver Gao, Eric J. Gonzales We propose an optimal credit-based pricing scheme for a single bottleneck with time-dependent demand and heterogeneous user preferences. Implementing such a strategy using a revenue neutral credit-based pricing scheme allows the value of the credits to be determined by the interaction between the users in the equilibrium condition of the market. The proposed strategy allocates the money raised from the “credit buyers” to subsidize the commutes of the “credit sellers” in order to incentivize the commuters to form a uniform distribution of arrival times over the peak period. As a result, designing a revenue neutral pricing strategy can raise public support by improving the social welfare of the users. Optimal Design Of The Large-scale Transit Systems In Urban Regions Using The Macroscopic Fundamental Diagram Lan Liu, Cornell University, Ithaca, NY, 14853, United States, ll745@cornell.edu, Mahyar Amirgholy, Mehrdad Shahabi, H. Oliver Gao In this research, we propose a continuum approximation model for optimizing the network structure (line spacing and stop spacing) and the operating characteristics (headway and fare) of the transit system by minimizing a linear combination of (1) the generalized cost that users experience in their trips, (2) the operating cost of the transit system for the agency, and (3) the external cost of the emission in region. The optimal design of the transit system can be derived by minimizing the total cost of the transportation system in three different network allocation scenarios: (i) mixed network (Bus), (ii) dedicated lanes (Bus Rapid Transit), and (iii) parallel networks (Metro). Tuesday Poster Competition Exhibit Hall Tuesday Poster Competition Competition Poster Session Dynamic Pricing And Demand Side Management In Smart Communities Vignesh Subramanian, University of South Florida, 5006, Bordeaux Village pl, Apt 201, Tampa, FL, 33617, United States, vigneshs@mail.usf.edu, Tapas K. Das Dynamic pricing will actively engage the electricity consumers, having an advanced metering infrastructure (AMI), in centralized demand side management (CDSM), a key to price stability and network reliability. We propose a quadratic binary programming model for a centralized controller to schedule the consumer load. The numerical result demonstrates how CDSM can lower the price peaks, reduce the reserve capacity of the generator and minimize the consumer’s hourly tariff. Multi-stage Stochastic Optimization For Considering Investment Risk In Conflict Prone Countries – A Case Study Of South Sudan Open source framework for energy system modeling - referred to as Tools for Energy Model Optimization and Analysis, Temoa - is employed to explore possible energy planning strategies for South Sudan. Stochastic optimization is utilized to explicitly consider the risk of conflict and the resultant damage to generators and transmission lines within the system. Because data related to both conflict probabilities and damage are subjected to deep uncertainty, we rely on sensitivity analysis to generate key insights. Results show that while large, centralized plants benefit from economies of scale, distributed solar photovoltaic are more resilient to conflict. Neha Satish Patankar, PhD Student, NC State University, 2366 Champion court, Raleigh, NC, 27606, United States, nspatank@ncsu.edu

Two-stage Methodology For Multiobjective Robust Decision Making With Application In Water-energy Planning Daniel Jornada, Texas A&M University, 2734 San Felipe Dr, College Station, TX, 77845, United States, djornada@tamu.edu, V.Jorge Leon The large number of compromise solutions to choose from a multiobjective program poses significant challenge for decision making. We formalize a two- stage optimization methodology to narrow the alternatives under consideration by introducing secondary robustness criteria to hedge against implementation uncertainties. A water-energy planning problem illustrates the significance of the methodology. Modeling And Maximizing Power For Wind Turbine Arrays Lucas Buccafusca, University of Illinois Urbana-Champaign, Urbana-Champaign, IL, United States, buccafus@illinois.edu This talk considers a specific application domain, that of wind turbine arrays, and explores the use of partitioning and control design to optimize energy extraction. Large wind turbine arrays, or wind farms, can be viewed as coupled networks, which present many problems when applying traditional optimization techniques. In our work, we apply heuristic techniques, exploiting the inherent symmetry found in wind turbine arrays, to obtain simplified models for large arrays. Using these simplified models we first consider a dynamic programming-like approach to maximize power extraction under the condition of uniform wind. Using A Private Marketplace To Build A Hybrid Workforce For IT Service Delivery Monica Johar, Associate Professor, University of North Carolina, 9201 University City Blvd, Friday 352 C, Charlotte, NC, 28223, United States, msjohar@uncc.edu, Su Dong, Ram Kumar The emergence of on-demand service marketplaces is a relatively new phenomenon. Technology is facilitating innovative work arrangements using an on-demand workforce. As the range of services available on such marketplaces increases, organizations could explore innovative uses of on-demand workers. Organizations can explore work arrangements that benefit from using a hybrid workforce that consists of full-time and on-demand workers. This paper addresses this interesting new work paradigm by presenting a mathematical programming model of service delivery that leverages in-house workers and on-demand marketplaces for service delivery. A New Class Of Measures For Independence Test With Its Application In Big Data Qingcong Yuan, PhD Candidate, University of Kentucky, 305 MDS Building, 725 Rose Street, Lexington, KY 40506, Lexington, KY, 40506, United States, qingcong.yuan@uky.edu, Xiangrong Yin We introduce a new class of measures for testing independence between two random vectors, using characteristic functions. By choosing a particular weight function in the class, we study a new index for measuring independence and its property. Sample versions and their asymptotic properties using different estimations are developed. We demonstrate the advantage of our methods via simulations and real data. In particular, we illustrate the effective use of our methods in big data analysis. A Time-series System To Predict Glucose Concentrations Based On Continuous Glucose Monitoring Lei LI, Beihang University, Beijing, 100191, China, lilei19940219@163.com, Yimeng Shi, Jun Yang, Xiaolei Xie The estimated prevalence of diabetes in Chinese adults in 2013 was 11.6%, which for the first time surpassed the U.S. In recent years, Continuous Glucose Monitoring (CGM) systems are developed to record the patient’s daily blood glucose level. Such systems provide us the real- time glucose level. Recently, researchers implemented AR or ARMA models on a small pool of CGM data to predict future glucose level. In this study, we developed a method by using adaptive Autoregressive Integrated Moving Average (ARIMA) model on a larger data- set rather than models with fixed-order, which is more practical and accurate as the order of the whole data is unknown before. Optimizing Screening Policies Inside A Food Production Facility Nicole T Lane, PhD Candidate, North Carolina A&T State University, 3511 Carrington Street, Greensboro, NC, 27407, United States, nicole.t.lane@gmail.com, Lauren Berrings Davis New legislation requires food production facilities to have a food safety plan including mitigation strategies to increase security. This research identifies an optimal set of implementable strategies. The two-stage stochastic model presented incorporates the need for production minimums and food safety constraints. The results of a numerical study show that for relatively low costs, the implementation of these policies ensures that products leaving the facility are safe for consumption.

300