Operations Research, Systems Engineering and Industrial Engineering Commons^™

Examination And Utilization Of Rare Features In Text Classification Of Injury Narratives, Hsin-Ying Huang

Open Access Dissertations

Thanks to the advances in computing and information technology, analyzing injury surveillance data with statistical machine learning methods has grown in popularity, complexity, and quality over recent years. During that same time, researchers have recognized the limitations of statistical text analysis with limited training data. In response to the two primary challenges for statistical text analysis, dimensionality reduction and sparse data, many studies have focused on improving machine learning algorithms. Less research has been done, though, to examine and improve statistical machine learning methods in text classification from a linguistic perspective.

This study addresses this research gap by examining the …

Impact Of Rfid Information-Sharing Coordination Over A Supply Chain With Reverse Logistics, Juan Jose Nativi Nicolau

Open Access Dissertations

Companies have adopted environmental practices such as reverse logistics over the past few decades. However, studies show that aligning partners inside the green supply chain can be a substantial problem. This lack of coordination can increase overall supply chain cost. Information technology such as Radio Frequency Identification (RFID) has the potential to enable decentralized supply chain coordinate their information. Even though there are research that address RFID on traditional supply chain, few researches address how to coordinate RFID information sharing in a green supply chain. We study, through simulation experiments, two types of RFID information-sharing coordination under different configurations related …

Adaptive Sampling Trust-Region Methods For Derivative-Based And Derivative-Free Simulation Optimization Problems, Sara Shashaani

Open Access Dissertations

We consider unconstrained optimization problems where only “stochastic” estimates of the objective function are observable as replicates from a Monte Carlo simulation oracle. In the first study we assume that the function gradients are directly observable through the Monte Carlo simulation. We propose ASTRO, which is an adaptive sampling based trust-region optimization method where a stochastic local model is constructed, optimized, and updated iteratively. ASTRO is a derivative-based algorithm and provides almost sure convergence to a first-order critical point with good practical performance. In the second study the Monte Carlo simulation is assumed to provide no direct observations of the …

Parametric Approaches To Fractional Programs: Analytical And Empirical Study, Chong Hyun Park

Open Access Dissertations

Fractional programming is used to model problems where the objective function is a ratio of functions. A parametric modeling approach provides effective technique for obtaining optimal solutions of these fractional programming problems. Although many heuristic algorithms have been proposed and assessed relative to each other, there are limited theoretical studies on the number of steps to obtain the solution. In this dissertation, I focus on the linear fractional combinatorial optimization problem, a special case of fractional programming where all functions in the objective function and constraints are linear and all variables are binary that model certain combinatorial structures. Two parametric …

Best Matching Processes In Distributed Systems, Mohsen Moghaddam

Open Access Dissertations

The growing complexity and dynamic behavior of modern manufacturing and service industries along with competitive and globalized markets have gradually transformed traditional centralized systems into distributed networks of e- (electronic) Systems. Emerging examples include e-Factories, virtual enterprises, smart farms, automated warehouses, and intelligent transportation systems. These (and similar) distributed systems, regardless of context and application, have a property in common: They all involve certain types of interactions (collaborative, competitive, or both) among their distributed individuals—from clusters of passive sensors and machines to complex networks of computers, intelligent robots, humans, and enterprises. Having this common property, such systems may encounter common …

Markov-Based Ranking Methods, Baback Vaziri

Open Access Dissertations

Ranking methods are an essential tool to help make decisions. This dissertation document examines different aspects of the theory and application of pairwise comparison ranking methods, specifically those that use Markov chains. First, a new method is developed to solve a traditional recruiting problem, and is shown to improve the predictive power of its ranking. Next, modifications are made to an existing method that theoretically improves the reliability, while maintaining the rank integrity. Last, a framework is developed that defines a fair and comprehensive ranking method, and several popular methods are evaluated in their ability to adhere to the said …