2015 Informs Annual Meeting

WB17

INFORMS Philadelphia – 2015

4 - Risk-averse Control of Diffusion Processes Jianing Yao, Doctoral Student, Rutgers University, 18M Reading Road, Edison, NJ, 08817, United States of America, jy346@scarletmail.rutgers.edu In this paper, we study the risk-averse control problem for diffusion processes. We make use of a forward-backward system of stochastic differential equations to evaluate a fixed policy and to formulate the optimal control problem. Weak formulation is established to facilitate the derivation of the risk-averse dynamic programming equation. We prove that the value function of the risk-averse control problem is a viscosity solution of a risk-averse analog of HJB equation. WB15 15-Franklin 5, Marriott Healthcare Informatics Sponsor: Optimization in Healthcare Sponsored Session Chair: Satyender Goel, Research Assistant Professor, Northwestern University, 633 St. Clair St, 20th floor, Chicago, Il, 60611, United States of America, s-goel@northwestern.edu 1 - Hep C Registry to Link, Track, Test and Treat Patients in Chicago Satyender Goel, Research Assistant Professor, Northwestern University, 633 St. Clair St, 20th Floor, Chicago, IL, 60611, United States of America, s-goel@northwestern.edu, Abel Kho Health information technology is the key to identify and track specific patient population cohorts and improve public health surveillance and follow-up. The CDC-funded Hepatitis C Community Alliance to Test and Treat (HepCCATT) technical infrastructure supports connections among multiple public and private partners including; community clinic sites, community and specialty pharmacies, public and private healthcare systems, city and state departments of public health, and labs. 2 - The Pragmatic Trial App Suite (PTAS): using Tablets to Engage Patients at the Point of Care Eliel Oliveira, Associate Director, LPHI, eoliveira@lphi.org, PTAS is an electronic medical record agnostic, tablet-based software system integrated in examination rooms designed to engage patients in research by electronically recruiting them in pragmatic research trials, providing targeted health information, and facilitating trial management. A pilot of PTAS is currently recruiting patients into a research network. The pilot began in March 2015 and has recruited 320 patients into HiOH (30%) and collected GLOBAL PROMIS surveys from 457 patients (40%). WB16 16-Franklin 6, Marriott Game Theory III Contributed Session Chair: Igor Kozeletskyi, University of Duisburg-Essen, Lotharstr. 65, Duisburg, 47057, Germany, igor.kozeletskyi@uni-due.de 1 - Finding a Nash Equilibrium in EPCC: An Application to Dynamic Electricity Markets Sébastien Debia, PhD Candidate, HEC Montréal, 10-6100 Av, Wilderton, Montréal, QC, Canada, sebastien.debia@hec.ca Equilibrium Problem with Complementarity Constraints (EPCC) formulation is widely used to analyze strategic interactions in electricity markets. However, each individual problem is not convex, and the collection of problems is not square. Hence a Nash Equilibrium is not proven to exist. This contribution aims at showing sufficient conditions to find numerically a Nash Equilibrium if it exists. We provide an application to dynamic electricity markets with hydro-power. 2 - Recycling Models of Power Batteries in Electric Vehicles Based on Product Life Cycle Theory Qihao Jin, Huazhong University of Science and Technology, School of Management, 1037 Luoyu Road, Wuhan, 430074, China, 390987833@qq.com, Zhixue Liu This paper studies impacts of different characteristics of electric vehicle power batteries recycling in four life cycles of the electric vehicle development on recycling models of power batteries. Closed-loop supply chains are established with the manufacturer, the Stackelberg leader, having three reverse channels: (1) it collects by itself, (2) it collects by a retailer, (3) it subcontracts the collection. Decisions of the manufacturer on recycling channels in four life cycles are considered. Iben Ricket, Elizabeth Nauman, Rebekah Angove, Kristin Layman, Lindsay Hendryx, Thomas Carton

3 - Cooperative Games with Multiple Objectives by Means of Bi-Allocation Games Igor Kozeletskyi, University of Duisburg-Essen, Lotharstr. 65, Duisburg, 47057, Germany, igor.kozeletskyi@uni-due.de, Alf Kimms, Ana Meca In this presentation a new class of non-transferable utility games is introduced. This class of games describes multi-objective cooperative situations where every player follows two objectives: his individual objective and a common objective for all players. For bi-allocation games we present a solution concept, defined as an extension of the Shapley NTU value, state that this value always exists and introduce a computation algorithm based on iterative search and multi-objective optimization. 4 - Easier Than We Thought – A Practical Scheme to Compute Pessimistic Bilevel Optimization Problems Bo Zeng, Assistant Professor, University of Pittsburgh, 3700 O’Hara Street, Pittsburgh, PA, 15261, United States of America, bzeng@pitt.edu We present a new computation scheme for pessimistic bilevel optimization problems (PBL), which so far does not have any computational methods generally applicable yet. We first develop a tight relaxation and then design a simple scheme to ensure a feasible and optimal solution to PBL. Then, we discuss using this scheme to compute linear PBL and several variants. We also provide numerical demonstrations on linear PBL problems. 5 - Claudico and the First Heads-up No-limit Texas Hold’em Man-machine Match Tuomas Sandholm, Professor, Carnegie Mellon University, 5000 Forbes Ave, Pittsburgh, PA, 15213, United States of America, sandholm@cs.cmu.edu, Sam Ganzfried, Noam Brown Claudico is the next-generation version of our Tartanian7 program, which won the 2014 Annual Computer Poker Competition in heads-up no-limit Texas Hold’em by beating all opponents with statistical significance. This talk will discuss Claudico and the “Brains vs. AI” match where four top-10 heads-up no- limit Texas Hold’em pros played 80,000 hands of poker against Claudico under controlled conditions for two weeks in 2015. WB17 17-Franklin 7, Marriott Network Optimization and its Applications Sponsor: Optimization/Network Optimization Sponsored Session Chair: Tachun Lin, Assistant Professor, Bradley University, 1501 W Bradley Ave, Peoria, IL, 61625, United States of America, djlin@fsmail.bradley.edu 1 - Routing Battery Electric Vehicles in Stochastic Networks Considering Vehicle Range Uncertainty Xing Wu, Assistant Professor, Lamar University, P.O. Box 10024, Cherry Engineering Building RM C-2032, Beaumont, TX, 77710, United States of America, xwu1@lamar.edu, Changzheng Liu The distance that a battery electric vehicle (BEV) can travel with a full charge varies greatly in the real driving environment. Based on the collected BEV travel data, this study assumes a stochastic effective battery range following a normal distribution. With such an assumption, this paper aims to develop the optimal routing policy for a BEV in a stochastic network, considering the uncertainty of both travel time and vehicle range. 2 - Selecting Inland Waterway Maintenance Projects Subject to Random Disruptions Khatereh Ahadi, University of Arkansas, 4207 Bell Engineering Center, Fayetteville, AR, 72701, United States of America, kahadi@uark.edu, Kelly Sullivan We study the problem of maintenance project selection for the inland waterway transportation network. Uncertainty arises from unpredictable natural/hydrologic conditions associated with shoaling. Thus, we propose a scenario-based stochastic programming model which considers project interdependence and multi- commodity flow between several supply and demand nodes. We model this problem as a mathematical program, develop solution approaches, and analyze computational results.

405