Informs 2015

71 / 552

INFORMS Philadelphia – 2015
69
2 - Reducing Healthcare Spending through Electronic Information
Exchange: Aligning Provider and Insurer
Idris Adjerid, Assistant Professor, Notre Dame, 358 Mendoza
College of Business, Notre Dame, United States of America,
Health information exchanges (HIEs) are entities that enable the electronic
sharing of patient information between disparate healthcare providers and other
stakeholders but little evidence that speaks to whether promised gains to quality
and efficiency have been realized. Leveraging a unique national panel dataset, we
find significant cost reductions in healthcare markets that have established
operational HIEs with an average reduction in spending of $139 per Medicare
beneficiary per year.
3 - The Spillover Effects of Health it Investments on Regional Health
Care Costs
Hilal Atasoy, Assistant Professor, Temple University, Fox School of
Business Alter Hall 445, Philadelphia, PA, United States of
America, , Pei-yu Chen, Kartik Ganju
Electronic medical records (EMR) are often presumed to reduce the significant
health care costs in the US. However, evidence on the impact of the EMR on costs
is mixed, leading to skepticism about their effectiveness. We argue that the
benefits EMR can go beyond the adopting hospital due to patient and physician
sharing. We find that EMR investments have significant spillover effects by
reducing costs of neighboring hospitals, which suggests that EMR can reduce
aggregate health care costs.
4 - Show Me the Way to Go Home: An Investigation of Ride Sharing
and Alcohol Related Vehiclar Homicide
Brad Greenwood, Assistant Professor, Temple University, 1810
North 13th Street, Philadelphia, PA, 19122, United States of
America, Sunil Wattal
We investigate how the entry of the driving service Uber affects the rate of
alcohol related motor vehicle homicides. Using a difference in difference
approach, the entry of Uber into markets between 2009 and 2013, results suggest
a significant drop in the rate of homicides during that time. Furthermore, results
suggest that not all services offered by Uber have the same effect, insofar as the
effect for the Uber Black car service is intermittent and manifests only in selective
locations.
SB11
11-Franklin 1, Marriott
Combinatorial Network Optimization
Sponsor: Optimization/Integer and Discrete Optimization
Sponsored Session
Chair: Cole Smith, Professor And Chair, Clemson University, Freeman
Hall, Clemson, SC, United States of America,
1 - On the Lagrangian Dual of the Maximum Quasi-clique Problem
Zhuqi Miao, Doctoral Candidate, Oklahoma State University,
322 Engineering North, Stillwater, OK, 74078, United States of
America, Baski Balasundaram
Quasi-clique is a useful model in cluster detection. The maximum quasi-clique
problem (MQCP) can be formulated as a binary quadratically constrained
program (BQCP). In this talk, we present new computational techniques based on
the Lagrangian dual of the BQCP formulation that can provide both good feasible
solutions and tight upper bounds for MQCP. We report results from our empirical
studies, which show that the proposed approaches outperform the leading
approaches for solving the MQCP.
2 - A Semi-Continuous Formulation for Maximizing Wireless Sensor
Network Lifetime
Rob Curry, Graduate Research Assistant, Clemson University,
Freeman Hall, Clemson, SC, 29634, United States of America,
, Cole Smith
A wireless sensor network consists of battery-powered sensors that collect and
transmit data from a set of targets. Receiving and transmitting data consumes
battery life; hence, multiple routes are employed to balance energy utilization and
maximize network lifetime. Also, a routing configuration may need to remain
stable for a minimum operating time if it is used at all. We formulate a semi-
continuous linear program for this problem, along with a branch-and-price
algorithm for its solution.
3 - A Branch and Price Algorithm for Solving the Hamiltonian
P-median Problem
Ahmed Marzouk, Texas A&M University, 3131, TAMU,
College Station, TX, 77843, United States of America,
Erick Moreno-centeno, Halit Uster
Given an undirected graph, G, the Hamiltonian p-median problem is to find p
cycles partitioning G with minimum cost. We present a Branch & Price algorithm
that solves instances up to 129 nodes (state-of-the art is 40-nodes). To solve the
pricing problem we developed 1) a new efficient algorithm to find the least cost
cycle in undirected graphs with arbitrary edge costs but no negative cycles; and 2)
an algorithm to find the most negative cycle in undirected graphs with arbitrary
edge costs.
4 - A Backward Sampling Framework for Interdiction Problems
with Fortification
Leonardo Lozano, PhD Student, Clemson University, 129
Freeman hall, Clemson, SC, 29634, United States of America,
Cole Smith
Three-stage sequential defender-attacker-defender problems are notoriously
difficult to optimize, especially when the third-stage (recourse) problem is
nonconvex. Our approach allows the third-stage problem to take any form. The
algorithm restricts the defender to select a recourse decision from a sample space,
and iteratively refines the sample to force convergence. We show the algorithm’s
effectiveness on various such problems, including games played over the TSP and
CLSP.
SB13
13-Franklin 3, Marriott
Recent Advances in Stochastic Integer Programming
Sponsor: Optimization/Optimization Under Uncertainty
Sponsored Session
Chair: Yongjia Song, Virginia Commonwealth University,
1015 Floyd Ave, 4140 Grace Harris Hall, Richmond, VA, 23284,
United States of America,
1 - On The Quantile Cut Closure of Chance-constrained Problems
Weijun Xie, Georgia Institute of Technology, School of ISYE,
755 Ferst Drive, NW, Atlanta, GA, 30332-0205, United States of
America, , Shabbir Ahmed
Quantile cuts for a chance-constrained problem (CCP) are projections of a subset
of the well-known mixing set inequalities onto the original problem space. We
study the closure of quantile cuts, and show that iterative application of the
closure recovers the convex hull of a CCP.
2 - A Decomposition Framework for Stochastic Mixed
Integer Programming
Suvrajeet Sen, Professor, University of Southern California,
University Park Campus, Los Angeles, CA, 90089, United States
of America,
We will summarize decomposition-based algorithms for SMIP problems. The key
to these algorithms are parametric cutting planes which are developed from one
of three principles: Gomory Cuts, Disjunctive Cuts, and Value Function Cuts.
These concepts lead to some of the most effective schemes for SMIP problems. In
a companion presentation, we will present computational results with a variety of
instances.
3 - A Polyhedral Study of Multistage Stochastic Unit
Commitment Polytope
Yongpei Guan, Professor, University of Florida, Weil 413,
Gainesville, FL, 32611, United States of America,
Jean-paul Watson, Kai Pan
In this study, we investigate a scenario-tree based multistage stochastic integer
programming formulation for the unit commitment problem under uncertainty.
By exploring its polyhedral structure, several families of strong valid inequalities
are generated. In particular, we obtain convex hull presentations for certain
special cases and facets for the general polytope. Finally, the computational results
verify the effectiveness of the proposed cutting planes.
SB13