2016 INFORMS Annual Meeting Program

WC21

INFORMS Nashville – 2016

WC18 106A-MCC

WC20 106C-MCC Understanding the US Index Futures Stock Market using Research Invited: Tutorial Invited Session Chair: William T. Ziemba, University of British Columbia, 1, Vancouver, BC, 2, Canada, wtzimi@mac.com 1 - Understanding The US Index Futures Stock Market Using Research William T. Ziemba, University of Bristish Columbia, Vancouver, BC, 2, Canada, wtzimi@mac.com I begin with five views or camps of beliefs concerning the US stock market. There are efficient markets where prices are correct except for transactions costs, risk premium where excess returns can be made only by bearing additional risk, efficient markets is hogwash, great investors exist but you cannot tell who they are in advance and the study of anomalies and other research. Edges arise from cash flows, institutional practices and behavioral biases. These include the turn of the year effect, the turn of the month effect, presidential election effects and mispriced options. I describe the effects and explain why they exist and then discuss their use in trading considering operational risks, the effect of volatility, prediction of stock market crashes, slippage, risk management, and optimal betting sizing. I won the 2015 futures trading contest of the Battle of the Quants in New York and have been able to obtain very good risk adjusted returns during July 2013 to May 2016 in the Alpha Z Futures Fund. WC21 107A-MCC Healthcare Operations and Capacity Planning Sponsored: Health Applications Sponsored Session Chair: Sukriye Nilay Argon, University of North Carolina, Hanes Hall CB# 3260, Chapel Hill, NC, 27599, United States, nilay@unc.edu Co-Chair: Serhan Ziya, University of North Carolina, 356 Hanes Hall, CB#3260, Chapel Hill, NC, 27599, United States, ziya@unc.edu 1 - Admission, Routing And Early Discharge Decisions In A Hospital Setting Lerzan E Ormeci, Koc University, lormeci@ku.edu.tr, Nermin Kurt, Amin Khoshkenar We consider the problem of bed management in a hospital. The patients stay at the hospital for a random length of time to recover after a surgery. Hence, the operation schedule has a long-term effect on the occupancy levels, which significantly affect the quality of care for the patients. To control the occupancy levels, hospital management has a number of options: 1) New patients may be rejected at high levels of occupancy, 2) Patients operated by a certain department may stay in the ward of another department, 3) Patients staying at the hospital may be discharged early. In this study, we analyze the effects of these decisions on the hospital performance by modeling the system via Markov decision processes. 2 - Ambulance Redeployment And Dispatching Under Uncertainty With Personnel Workload Limitations Shakiba Enayati, North Carolina State University, Raleigh, NC, United States, senayat@ncsu.edu, Osman Ozaltin, Maria Esther Mayorga Emergency Medical Services managers are concerned with providing maximum possible coverage in a service area. Redeployment refers to dynamic relocation of idle ambulances to compensate for the coverage loss due to busy ambulances. Unsystematic redeployments, however, impose superfluous workload on the personnel. We propose a two-stage stochastic programming model to redeploy and dispatch ambulances to maximize expected coverage. Our model includes personnel workload restrictions in a shift. We develop a decomposition algorithm to determine an upper bound and apply a branch and bound algorithm to find the optimal solution. We evaluate the performance using a largescale real dataset. 3 - An Optimization Approach For Coordinating Clinic And Surgery Appointments To Meet Access Delay Service Levels Esmaeil Keyvanshokooh, University of Michigan, Ann Arbor, Ann Arbor, MI, United States, keyvan@umich.edu, Mark P. Van Oyen Providing timely access to surgery is crucial for patients with high acuity diseases like cancer. In this paper, we present an optimization approach for coordinating clinic and surgery appointments to meet access delay service levels in Colorectal surgery (CRS). The methodology is applied to historical patient data for CRS to show its better performance than the current scheduling policy.

DMA Healthcare Contributed Session Chair: Kaiye Yu, Tsinghua University, Room 615,Shunde Building, Tsinghua University, Beijing, BJ100084, China, yky15@mails.tsinghua.edu.cn 1 - Optimal Reimbursement Schemes For Maternity Care Safety And Quality Cheng Zhu, McGill University, 701-801 Sherbrooke Est, Montreal, QC, H2L 0B7, Canada, cheng.zhu@mail.mcgill.ca, Beste Kucukyazici The amount of unnecessary C-Sections, which expose proven higher postpartum complications of mothers and newborns as well as heavy economic burden, has been increasing constantly and this growth raises great concerns for the policy makers. This research focuses on optimizing payment mechanisms to reimburse obstetricians, in order to reduce unnecessary C-sections while retain it for those who need it, resulting in enhanced birth quality with alleviated economic burden for overall health care system. The optimal reimbursement schemes are further verified empirically with large datasets. 2 - Using Policy Flight Simulator To Evaluate Scalability Of Evidence Based Practices Zhongyuan Yu, Research Assistant Professor, Stevens Institute of Technology, Hoboken, NJ, 07030, United States, zyu7@stevens.edu, William B Rouse, Michael Pennock, Kara Pepe Numerous evidence-based practices have demonstrated reduced medical costs, improved patient experiences and better quality of life. However, concern from stakeholders in health system seems to be holding back adoption. Policy flight simulator is proposed to find out what affects scalability of these practices, and assess what needs to be adjusted in order to increase confidence of senior administrators to expand these practices. Policy flight simulators fuse aspects of traditional scientific analysis, engineering, social science, and visualization to provide decision makers an immersive experience to increase comfort level with simulation-based data-driven decision making processes. 4 - Modeling And Assessment Of The Risk Of Colorectal Polyp Kaiye Yu, Tsinghua University, Room 615,Shunde Building, Tsinghua University, Beijing, BJ100084, China, yky15@mails.tsinghua.edu.cn, Jie Xing, Wenying Zhou, Shutian Zhang, Xiaolei Xie, Nan Kong Most colorectal cancer (CRC) arise from polyps. We identify colorectal polyps risk factors and develop risk stratification model using machine learning approaches. The individualized risk assessment tool could offer decision support to both clinicians and patients. WC19 106B-MCC Open-Source Tools for Operations Research Sponsored: Computing Sponsored Session Chair: Matthew J Saltzman, Clemson University, Clemson University, Clemson, SC, 29634, United States, mjs@clemson.edu 1 - New Developments In Pyomo William E Hart, Sandia National Laboratory, wehart@sandia.gov Pyomo is a Python-based open-source software package that supports a diverse set of optimization capabilities for formulating, solving, and analyzing optimization models. In this presentation, we describe new capabilities in Pyomo, including support for new versions of Python, installation with conda, and updates for modeling capabilities (bilevel, sp, dae, etc). This talk will also highlight new documentation resources for users. 2 - Jmarkov: An Integrated Java Framework For Stochastic Modeling Daniel F Silva, Georgia Institute of Technology, Atlanta, GA, United States, dfsi3@gatech.edu, Julio C Goez, Juan F Perez, German Riano jMarkov allows users to define stochastic models from the basic rules underlying the system dynamics and then solve the models to obtain performance measures. It is composed of four modules: (i) the main module supports Markov Chain models with a finite state space; (ii) jQBD enables the modeling of Quasi-Birth- and-Death processes; (iii) jMDP offers the capabilities to model and solve Markov Decision Processes; and (iv) jPhase supports the manipulation and inclusion of phase-type variables. In addition, jMarkov is highly extensible, allowing the users to introduce new modeling abstractions and solvers. In this talk we give an overview of jMarkov, as well as some examples.

433