Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

SA72

responding to an invitation to writing the report. Panelists

n SA72 West Bldg 211A

Dashi Singham, Naval Postgraduate School, 1411 Cunningham Road, Operations Research Department, Monterey, CA, 93943, United States Robert Shumsky, Professor, Tuck School, Dartmouth, Tuck School of Business, 100 Tuck Hall, Hanover, NH, 03755, United States Burak Eksioglu, Clemson University, Department of Industrial Engineering, 272 Freeman Hall, Clemson, SC, 29634, United States n SA74 West Bldg 212A Recent Algorithmic Advances on Multi-objective Integer Programming Sponsored: Multiple Criteria Decision Making Sponsored Session Chair: Hadi Charkhgard, University of South Florida, Tampa, FL, 33620-5350, United States 1 - OOES.jl: A Julia Package for Optimizing a Linear Function Over the Set of Efficient Solutions for Bi-objective Mixed Integer Linear Programming Alvaro Sierra-Altamiranda, University of South Florida, Tampa, FL, United States, Hadi Charkhgard We present `OOES.jl’, a flexible package for optimizing a linear function over the set of efficient solutions for bi-objective mixed integer linear programs. The proposed framework extends our recent algorithm, by adding two main characteristics: (1) The package allows using any single-objective mixed integer programming solver supported by `MathProgBase.jl’; (2) The algorithm works under multiprocessing environments by exploiting parallelization. An extensive computational study evaluates the performance of the package versus previous C++ implementation, the performance of parallelization and provides a comparison between different solvers available in the market. 2 - A Branch-and-Bound Algorithm for a Class of Mixed Integer Linear Maximum Multiplicative Programs: A Multi-objective Optimization Approach Payman Ghasemi Saghand, FL, United States, Hadi Charkhgard, Changhyun Kwon We present a linear programming based branch-and-bound algorithm for a class of mixed integer optimization problems with a bi-linear objective function and linear constraints. These problems can be viewed as a special case of optimization over the efficient set. It is known that without integer variables, such problems can be transformed into a Second-Order Cone Program (SOCP) and be solved by CPLEX SOCP. Also, such problems can be solved faster by a linear programming based algorithm. In this study, we embed that algorithm in an effective branch- and-bound framework to solve mixed integer instances. An extensive computational study shows that the proposed algorithm outperforms CPLEX SOCP solver. 3 - MSEA-1.0: A Multi-Stage Exact Algorithm for Bi-objective Pure Integer Linear Programming in Julia Hadi Charkhgard, PhD, The University of South Florida, Tampa, FL, United States, Aritra Pal We present a new exact method for bi-objective pure integer linear programming, the so-called Multi-Stage Exact Algorithm (MSEA). The method combines several existing exact and approximate algorithms in the literature to compute the entire nondominated frontier of any bi-objective pure integer linear program. Each algorithm available in MSEA has multiple versions in the literature. Hence, the main contribution of our research is developing a united framework for all these versions in MSEA. The package supports execution on multiple processors and users (if interested) can easily customize the package for their specificc problems. 4 - Enabling Energy Storage Sharing among Multiple Independent Users: A Multi-objective Optimization Approach Rui Dai, University of South Florida, Tampa, FL, United States, Hadi Charkhgard Energy storage sharing should be coordinated in a proper way to ensure the fairness and efficiency when allowing users exchange their stored energy. To address this challenge, this work proposes a multi-objective optimization based sharing strategy to operate an energy storage system shared by multiple independent users. In order to guarantee the fairness in the exchange of stored energy, the payoff for the user transferring their stored energy to other users is calculated. This strategy is formulated as a Nash bargaining problem, and solved through a multi-objective optimization approach. In addition, piecewise McCormick relaxation is adopted to linearize the existing bilinear terms.

Healthcare Delivery Policy and Implementation Sponsored: Manufacturing & Service Oper Mgmt/Healthcare Operations Sponsored Session Chair: Shima Nassiri, University of Michigan, Ann Arbor, MI, 48109, United States 1 - Outcomes-based Reimbursement Policies for Chronic Care Pathways Sasa Zorc, INSEAD, 1 Ayer Rajah Avenue, PhD Office, Singapore, 138676, Singapore, Stephen E. Chick, Sameer Hasija We develop an outcomes-based model of contracting in care for chronic patients, using data from United Kingdom’s NHS. The government contracts with healthcare providers in effort to maximise population health minus the cost. We consider the decision of whether to contract with individual healthcare providers or groups of such providers, as well as which contract type to use. Individual contracts fail to provide the desired incentives if providers under such contracts cooperate (collusion), however so do group contracts if group members fail to coordinate (free-riding). We demonstrate that individual outcomes-adjusted capitation contracts are the most resistant to these adverse effects. 2 - Predicting with Proxies: Improving Medical Risk Scores Hamsa Sridhar Bastani, Wharton School, Philadelphia, PA, United States Risk scores are often used by providers to target interventions, and by payers to estimate costs. However, such scores may not transfer well from one healthcare provider to another due to differences in physician behavior, medical coding practices, and patient risk factors. On the other hand, developing a different risk score for each provider may be infeasible due to data scarcity. We present a new technique for adapting an existing risk score to a new provider by learning important biases specific to that provider. We evaluate our technique on a diabetes prediction task and demonstrate both improved predictive performance as well as insights into underlying differences between providers. 3 - Outcome-based Pricing for Pharmaceuticals Elodie Adida, University of California - Riverside, School of Business Administration, 900 University Avenue, Riverside, CA, 92521, United States We study the effect of outcome-based pricing for pharmaceuticals. Under this payment scheme, the pharmaceutical firm is paid only when the drug treatment achieves a pre-specified goal. We consider heterogenous, price-sensitive, risk- averse patients, a payer, and a pharmaceutical firm producing a drug with uncertain effectiveness. We find that outcome-based pricing is unlikely to solve the issues of high drug prices and high payer expenditures. However, supplementing outcome-based pricing with a transfer payment between firm and payer can make patients, payer and firm better off than under uniform pricing. 4 - The Price of Simplicity in Personalized Prostate Cancer Screening Strategies John Silberholz, University of Michigan, Ross School of Business Patient preferences for different health states can significantly impact their best course of action in screening for diseases like prostate cancer. Patients with a relatively small disutility for treatment side effects might benefit from aggressive screening, while others might benefit from less aggressive screening or no screening at all. In this work, we use a mathematical model to quantify the benefit of fully personalized prostate cancer screening versus a one-size-fits-all strategy. Further, we identify simpler, more interpretable personalized screening strategies that could be easier to implement in practice, and we quantify the price of this simplicity in strategies. n SA73 West Bldg 211B JFIG Panel Discussion: Best Practices in Reviewing Papers Sponsored: Junior Faculty JFIG Sponsored Session Chair: Canan Gunes Corlu, Boston University, Boston, MA, 02215, United States Co-Chair: Gokce Palak, Shenandoah University, Winchester, VA, 22601, United States 1 - JFIG Panel Discussion: Best Practices in Reviewing Papers Canan Gunes Corlu, Boston University, 808 Commonwealth Avenue, Boston, MA, 02215, United States Past and current associate editors will provide advice about reviewing papers from

28