Results :: Search Electronic Theses and Dissertations (ETDs)

Genepart Algorithm, Clustering and Feature Selection for DNA Micro-Array Data: Author: Zhang, Weihua; Date: 2004-12-15; Program: Computer Science; Abstract: This paper provides the theoretical analysis of a new clustering and feature selection algorithm for the DNA micro-array data. This algorithm utilizes a branch and bound algorithm as the basic tool to quickly generate the optimal tissue sample partitions and select the gene subset which contributes the most to certain sample partition, it also combines the statistical probability method to identify important genes that have meaningful biological relationships to the classification or clustering problem. The proposed method combines feature selection and clustering processes and can be applied ...; Download File | View full details

Map Labeling with Circles: Author: Jiang, Minghui; Date: 2005-05-15; Program: Computer Science; Abstract: We study two geometric optimization problems motivated by cartographic applications: Map Labeling with Uniform Circles (MLUC) and Map Labeling with Uniform Circle Pairs (MLUCP). We show that the decision problems of both MLUC and MLUCP are NP-hard, and that the related optimization problems for maximizing the label sizes are NP-hard to approximate within factor 1.0349. We design approximation algorithms with constant performance guarantees for the two problems: for MLUC, we present a (3 + ε)-approximation and a (2.98 + ε)-approximation; for MLUCP, a (1.5+�µ...; Download File | View full details

Project Task: Member Assignment Using Design Structure Matrix and Genetic Algorithm in Concurrent Engineering Project Management: Author: Mazur, Lukasz Maciej; Date: 2005-05-15; Program: Industrial and Management Engineering; Abstract: In concurrent engineering, project tasks generally require the establishment of multifunctional teams to simultaneously consider various activities throughout the entire product life cycle. Team members from different functional departments of the company interact in every phase of development activities to design products and processes concurrently. This concurrent strategy increases the complexity of product development and design processes and makes teams difficult to organize. Without effective task coordination and team organization, the lack of communication and cooperation among team me...; Download File | View full details

Generic Properties of the Infinite Population Genetic Algorithm: Author: Hayes, Christina Savannah Maria; Date: 2006-08-15; Program: Mathematics; Abstract: The infinite population model for the genetic algorithm, where the iteration of the genetic algorithm corresponds to an iteration of a map G, is a discrete dynamical system. The map G is a composition of a selection operator and a mixing operator, where the latter models the effects of both mutation and crossover. This dissertation examines the finiteness and hyperbolicity of fixed points of this model. For a typical mixing operator, the fixed point set of G is finite and all fixed points are hyperbolic....; Download File | View full details

Animating the Conversion of Nondeterministic Finite State Automata to Deterministic Finite State Automata: Author: Merryman, William Patrick; Date: 2007-05-15; Program: Computer Science; Abstract: Many students may find the conversion of non-deterministic finite state automata into deterministic finite state automata to be difficult. Since standard computers are deterministic by nature, it is beneficial to understand how to convert nondeterministic finite state automata into equivalent deterministic versions. This conversion process is often inadequately presented in traditional textbooks, though, as static presentations rarely capture the dynamics of the process. This thesis provides a Java applet that will help students better understand the conversion, and may help teachers to better...; Download File | View full details

Apriori approach to graph-based clustering of text documents: Author: Hossain, Mahmud Shahriar; Date: 2008-05-15; Program: Computer Science; Abstract: This thesis report introduces a new technique of document clustering based on frequent senses. The developed system, named GDClust (Graph-Based Document Clustering) [1], works with frequent senses rather than dealing with frequent keywords used in traditional text mining techniques. GDClust presents text documents as hierarchical document-graphs and uses an Apriori paradigm to find the frequent subgraphs, which reflect frequent senses. Discovered frequent subgraphs are then utilized to generate accurate sense-based document clusters. We propose a novel multilevel Gaussian minimum support strat...; Download File | View full details