Surgery Scheduling with Recovery Resources,
Industrial and Operations Engineering, University of Michigan
Surgery scheduling is complicated by the post-anesthesia care unit, the typical recovery resource. Based on collaboration with a hospital, we present a novel, fast 2-phase heuristic that considers both surgery and recovery resources. We show that each phase of the heuristic has a tight provable worst-case performance bound. Moreover, the heuristic performs well compared to optimization based methods when evaluated under uncertainty using a discrete event simulation model.
A Branch-and-Cut Algorithm for the Bi-level Clique Interdiction Problem
I introduce an algorithm to solve the current formulation of the bilevel clique interdiction problem. The problem defines a defender who attempts to protect cliques from removal by an attacker. I use a branch and cut approach to solve the proposed problem and give preliminary results. This algorithm is expected to be usable on any social network, including terrorist cells or marketing strategies.
Modeling Individual Consumer Food Contamination Progression
North Carolina A&T State University
Food-borne illness affects nearly 48 million individuals a year resulting in hospitalizations and deaths. United States public health departments reported that 1,527 food outbreaks occurred between 2009 and 2010 of which 7.8% resulted in deaths. The purpose of this research is to develop models that will help to quantify consumer morbidity, consider the impact of various characteristics on the consumer, spread of contamination and consider interventions.
Methodology for Developing Chemical Portfolios from Full-Text
Department of Industrial and Manufacturing Engineering, The Pennsylvania State University
Chemical portfolios, the collection of interacting chemicals within a body of literature, are used to improve efficiency in manual learning. A methodology is developed to automatically create chemical portfolios based on connectivity. A modified graph diversity heuristic measure is developed. Results are presented from over 700,000 full-text journal files. Two novel findings are made: a relationship between the portfolio and corpus and an indirect link between diabetes and honeybee venom.
Integrated make-pack-route scheduling for fresh agri-food online retailing in China
Dalian University of Technology, China
Fresh agri-food sold via make-to-order farm-to-home online retailing undertakes a make-pack-route process at the farm’s distribution center. Due to large variety of raw produce, high cost of intermediate storage, flexible combination of finished products, and demand for timely delivery, integrated scheduling is needed. We thus introduce a make-pack-route optimization model and propose a local search based heuristic method. We report case studies based on real-world business practice in China.
Patient-Centered Barriers to Implementation of Remote Health Technologies for Diabetes Care: A Systematic Review
Department of Industrial & Systems Engineering, Texas A&M University
The objective of this systematic reviews was to identify barriers clinicians and patients face when using remote health technology for the management of Type 2 Diabetes. Six databases were searched for relevant studies published from Jan. 2010 - June 2015. The 58 articles selected for inclusion implemented remote health technology for disease management of adult patients with Type 2 Diabetes in the U.S. Data was collected andanalyzed for several categories including the terminology, outcome measures, retention rate, types of technology, and barriers reported in each study.
A framework for modeling the DCIS progression into IBC
Shadi Hassani Goodarzi
North Carolina State University
Ductal Carcinoma In Situ (DCIS) is arguably a direct precursor of Invasive Breast Cancer (IBC). Approximately 14%–53% of DCIS turn into IBC, after long follow-up periods ranging from 18 to 30 years (Bayraktar 2013i). So about 47%-86% of the DCIS cases are over diagnosed and as a result, treatment can only cause harm for these patients. This framework will allow us to study the progression of DCIS into IBC more clearly and as a result, aiding both patients and doctors in decision making.
Model Based Approach to Multi-Objective Optimization
Joshua Q Hale
H. Milton Stewart School of Industrial and Systems Engineering, Georgia Institute of Technology
We develop a model-based algorithm for the optimization of multiple objective functions that can only be assessed through black-box evaluation. The algorithm iteratively generates candidate solutions from a mixture distribution over the solution space and updates the mixture distribution based on the sampled solutions' domination count such that the future search is biased towards the set of Pareto optimal solutions.
Appointment Scheduling with No-Shows and Advance Cancellations
Joseph M. Katz Graduate School of Business, University of Pittsburgh
Appointment no-shows and cancellations can be disruptive to clinic operations. Scheduling strategies such as overbooking or overtime slot assignments can assist with mitigating these disruptions. We propose a scheduling model that accounts for both no-show and cancellation rates, and show properties of optimal scheduling models under specific conditions.
A Bayesian Approach to Value of Information: The Value of Waiting during a Trial of Labor
North Carolina State University
In the U.S. almost 1 in 3 births is a cesarean section (CS). Due to the implications CS may have on subsequent pregnancies and the health system as a whole, an evaluation of the first CS is of major importance. We present a Bayesian decision model that determines under what conditions it is appropriate to decide CS or trial of labor based on the belief that a patient is failure-to-progress by evaluating the tradeoff between prolonging labor and making an incorrect decision about patient type.
MDP Policy Approximation Via Poisson Regression
Wesley J. Marrero
Industrial and Operations Engineering, University of Michigan
Markov decision process (MDP) models are powerful tools which enable the derivation of optimal treatment policies, but may incur long computational times and decision rules which are challenging to interpret by physicians. To reduce complexity and enhance interpretability, we study how Poisson regression may be used to approximate optimal hypertension treatment policies derived by a MDP for maximizing a patient’s expected discounted quality-adjusted life years.
Effects of market segmentation on vaccine affordability in a centralized vaccine market
Industrial and Systems Engineering, Rochester Institute of Technology
This study analyzes a centralized vaccine market using mathematical optimization and simulation to understand the changes in affordability for low income countries as the number of market segments varies. Considering that the willingness to pay for vaccines is stochastic, our preliminary results show that there is a tradeoff between affordability and manufacturers' profit that leads to an ideal market segmentation.
Modeling Multiple Chronic Conditions
Industrial & Systems Engineering, North Carolina State University
Comorbidity is the presence of two or more chronic conditions in an individual. A 2012 CDC report found that 1 in 4 adults live with comorbidity, contributing to almost two-thirds of healthcare spending. Diabetes, a chronic condition by itself, is associated with several other chronic comorbidities. The goal of this research is to model multiple chronic conditions in hospitalized patients with diabetes so as to assess gender/racial disparities in the impact of comorbidity on discharge outcomes.
Will the cubic L1 spline fits well preserve a convex shape?
Edward P. Fitts Department of Industrial and Systems Engineering, North Carolina State University
Cubic L1 spline fits have shown some favorable shape-preserving property for geometric data. We consider the basic shape of a convex corner in a given window. Both analytical and numerical results have shown that the spline fit is convex if and only if the breaking point is in the middle third interval. It can be used as a reference to predict the approximating results of the spline fit in various circumstances.
Modeling the Spread of Measles with Pockets of Low Vaccination Coverage Rates
North Carolina A&T
While the national and state vaccination coverage rates of measles remain high, recent news has shown several outbreaks, which may be due to pockets of low vaccination coverage. We study this phenomenon by modeling vaccination rates in communities in Ohio. We have developed a spatial S-V-I-R (Susceptible-Vaccinated-Infected-Recovered) model.
Willingness-to-Pay (WTP) models for water in Post-Disaster Environments
Diana G. Ramirez-Rios
Department of Civil and Environmental Engineering , Rensselaer Polytechnic Institute
This paper focuses on willingness-to-pay (WTP) models in post-disaster environments where the distribution of critical supplies are severely compromised and the conditions are most critical. Econometric models were estimated to determine an economical valuation of WTP for water that could be included in post-disaster humanitarian logistics formulations. The models show a non-linear relation between willingness to pay, deprivation time, the expected time to normality and the socio-economic background of the individuals.
What Drives Economic Contagion? Findings from a Borrower-Lender Game
Jonathan W. Welburn
Department of Industrial & Systems Engineering, University of Wisconsin - Madison
The recent Eurozone debt crisis has demonstrated a significant source of risk to economic stability; adverse economic events in one country can quickly spread across countries and regions. While other post-crisis spillover effects occur over extended periods of time, this spread of crises, a phenomenon known as contagion, results in effects that occur in the short-term. The process by which contagion occurs can follow from cascading shocks spread through trade and debt channels. Alternatively, the same effect can be produced by common-cause shocks where multiple countries experience similar adverse shocks simultaneously. Our aim is to elucidate the key driver of contagion, whether it be debt, trade, the combined effect of debt and trade, or a common cause. We adopt an interdisciplinary approach by combining methods from engineering risk analysis with traditional economic modeling. We present a within-period sequential-move game with two borrower countries and a single lender to model the process of contagion immediately following a crisis. Each borrower can receive adverse independent and common-cause shocks, which can then spread through debt and/or trade channels. We model the effect of specific contagion channels through subgames--a debt model, a trade model, and a model with both debt and trade (to highlight their interaction). Furthermore, we present computational results calibrated to the 2010 Eurozone crisis. We discuss how contagion could occur through each channel, demonstrate that lender beliefs can drive contagion even in the absence of cascading shocks, and give recommendations on how future crises could be managed.
A Predictive Model of Smoking Prevalence Based on Individual Dynamics
Operations Research Graduate Program, North Carolina State University
We constructed a Markov model for individual adults’ smoking dynamics. Transition matrices were developed for the two age groups: 18-34 and 35+, respectively. A discrete event simulation model was built to forecast smoking prevalence through 2020. Then, we showed the applications of the simulation model.