A Simulation and Optimization Approach for Scheduling Chemotherapy Appointments under Uncertainty
Michelle M. Alvarado
Texas A&M University
The cost and demand of chemotherapy treatments in the United States is on the rise. Scheduling chemotherapy appointments is challenging because patients can experience adverse reactions during treatment that result in uncertain appointment durations and nurse acuity levels. We use a discrete-event simulation to model clinic operations and a stochastic integer programming model to schedule patients and resources. The methodology is evaluated using patient data from a Texas oncology clinic.
Managing rail-truck intermodal transportation for hazardous materials with congestion considerations
Ghazal Assadipour
Faculty of Business Administration
Memorial University
This research analyzes the problem of capacity planning and routing of regular and hazardous materials in a rail-truck intermodal network when the demand for transportation is stochastic. The novel feature of the suggested model is the consideration of congestion, as a source of exposure and delay, when making equipment acquisition and routing decisions. We illustrate the application of the model using a real problem instance based on the intermodal service chain of Norfolk Southern in US.
Generalizations of Continuous Mixing Set: Extended Formulations and Facets
Manish Bansal
Department of Industrial and Systems Engineering
Texas A&M University
We present our progress in studying some interesting generalizations of the well-known continuous mixing set. We develop extended formulations and new facet-defining inequalities for these sets. We also present exact separation algorithm for the developed cuts and computationally evaluate the effectiveness of these cuts for multi-module lot-sizing problems with(out) backlogging.
Evaluating Consumer Morbidity from Food Supply Chain Contamination
Jessye Bemley
North Carolina A&T State University
This research aims to compare the vulnerabilities of three food distribution channels, in terms of consumer morbidity from consumption of an intentionally contaminated food item. A discrete time Markov Chain with rewards is presented to quantify the risk of exposure and the effect of a contaminant dose on the consumer.
Optimal Patient-centered Resuscitation for Hospitalized Patients Using Electronic Medical Records
Muge Capan
Edward P. Fitts Industrial and Systems Engineering Department
North Carolina State University
Every day caregivers make decisions regarding the recovery and stabilization (i.e., resuscitation) of patients in case of deterioration in health. Timely response, such as initiating a Rapid Response Team (RRT), is critical to prevent further decline and improve patient outcomes. We develop semi-Markov Decision Process models to identify patient specific RRT activation rules using electronic medical records. The numerical results provide insight into the control-limit structure of the optimal policies.
A Mixed-Integer Optimization Model for the Economic and Environmental Analysis of Biomass Production
Halil Cobuloglu
Department of Industrial and Manufacturing Engineering
Wichita State University
In this study, we propose a mixed-integer optimization model, which defines the best decision strategies for biomass production regarding seeding time and methods, land allocation, harvesting amount and time under limited budget while considering environmental impacts. Results are presented for switchgrass cultivation as a source of biomass in the State of Kansas.
Chance-Constrained Models and Approaches for Stochastic Operating Room Allocation and Scheduling
Yan Deng
Department of Industrial & Operations Engineering
University of Michigan
We consider a stochastic operating room (OR) planning problem with uncertain surgery durations, which minimizes operating cost through OR opening, surgery allocation and scheduling decisions. Decisions are connected by MIP constraints to ensure validity of the plan, and by chance constraints to enforce quality requirements. We use decomposition and a strengthened master problem (SMP) to avoid curse of dimensionality. We also develop and lift inequalities for SMP to improve computational efficacy.
A Model for Forecasting Binary Outcomes Resulting from Human Behavior
Shannon Harris
Joseph M. Katz Graduate School of Business
University of Pittsburgh
We consider forecasting the next outcome of a binary human behavior process, based only on its historical values. Our model has three components that capture critical aspects of the underlying behavioral process. We illustrate our approach using data from patients’ attendance and non-attendance at VA outpatient clinics.
Decision Analytic and Bayesian Models for VBAC Decision Making Considering Mother and Child Outcomes
Karen Hicklin
North Carolina State University
In the U.S. the rate of cesarean sections (C-sections) was 32.8% in 2011 making it the most common major surgical procedure. C-sections are associated with both maternal and neonatal risks and can cause complications in subsequent pregnancies. We model the trade-off between C-section and trial of labor considering both mother and child outcomes as a function of disutility associated with mode of delivery for a women with a prior C-section. We also develop a Bayesian decision model to analyze how long a woman should remain in labor before deciding she needs a C-section.
When is Radionuclide Bone Scan Needed for the Baseline Staging of Newly Diagnosed Prostrate Cancer?
Selin Merdan
Department of Industrial and Operations Engineering
University of Michigan
Radionuclide bone scans are often performed in the staging evaluation of patients with newly diagnosed prostate cancer, but not all patients are at the same risk of developing bone metastases. This poster will present a study of methods to predict bone scan findings using prostate-specific antigen levels and other clinical parameters, and to identify the subgroup of patients for whom a baseline staging bone scan may be safely omitted without compromising the ability to detect metastatic disease.
An Integrated Response and Recovery Model for Disaster Relief Operations Support
Kasin Ransikarbum
Department of Industrial Engineering
Clemson University
While logistics research recently has placed increased focus on disaster relief operations, few studies have examined the response and recovery phases in post disaster operations. We present an integrated network optimization model for making strategic decisions in both the supply distribution and network restoration domains. We discuss equity-based solutions for constrained budget problems to provide decision makers with candidate restoration and distribution plans.
A Fairness in Dispatching Emergency Vehicles under Multi-tiered Response
Kanchala Sudtachati
Department of Industrial Engineering
Clemson University
We investigate the performance of two alternative dispatching policies in EMS systems for double unit dispatches, considering three priorities levels. The simulation model determines how to dispatch ambulances for priority2 calls considering fairness constraints. The objective is to maximize the expected survival rate. Considering fairness could improve the effectiveness of EMS systems.
Stochastic Integer Programming Models for Air traffic flow management
Saravanan Venkatachalam
Industrial and Systems Engineering Department,
Texas A&M University
Our work on Stochastic Integer Programming (SIP) Model for Air traffic flow management (ATFM) deals with the strategic problem of resolving imbalances between air traffic system demand and capacities through the definition of flight control actions. We developed a decomposition algorithm constituting L-Shape method, Cut generation procedure to generate Fenchel Cuts and techniques for upper bounding for faster convergence. We present the details of implementation and the experience of computational study for the instances based out of European air traffic.
A Game Theoretic Model of Financial Crises
Jonathan W. Welburn
University of Wisconsin –Madison
Department of Industrial & Systems Engineering
Global financial crises have revealed the systemic risk posed by economic contagion. We formulate a game between countries, central banks, banks, firms, households, and financial inter-governmental organizations to model the dynamics of borrowers and lenders and endogenize financial frictions. We model strategic choices, determine equilibrium solutions, and simulate the impacts of random shocks. Our conclusions enhance the understanding of global economic risk.
The Minimal k-core Problem for Modeling k-assemblies
Cynthia Wood
Rice University
I present a backtracking algorithm to find all minimal k-cores of a given undirected graph, which belongs to the class of NP-hard problems. The proposed method is a modification of the Bron and Kerbosch algorithm for finding all cliques of an undirected graph. The minimal k-core problem has applications in the area of neuroscience. For example, in the study of associative memory, a cell assembly is a group of neurons that are strongly connected and represent a “oncept” of our knowledge. This group is wired in a specific manner such that only a fraction of its neurons will excite the entire assembly. Recent studies have linked the concept of a particular type of cell assembly called k-assembly to the closure of a minimal k-core. Therefore, the proposed method puts us a step closer to test its mathematical definition.