Thesis Details

Thesis Title: Improving Product and Process Quality with Data Mining
Thesis Author: Beth Anderson
Abstract: Information technology has become a priority in many businesses today. Understanding the data collected in information systems has allowed companies to stay competitive in different industries. Data mining is a set of statistical techniques for discovering previously unknown trends and patterns in large datasets. Data mining is becoming a significant tool in todays competitive business world and helping business leaders to make more informed decisions. The purpose of this research has been to understand the yarn manufacturing process, the data collected, and the different data collection systems in order to determine the potential application of data mining in cotton yarn spinning. Plant interviews and a case study with a cotton open-end spinning plant were conducted to understand how data mining could be used within yarn manufacturing. Data was collected from the case study plant and analyses were performed to determine relationships and trends. This research provides an overview of the data collection requirements for textile spinning and the different data elements collected throughout the spinning process. A model of collection points and data elements is presented. The different data collection systems used for monitoring these elements are discussed as well as the quality of data being collected. Analyses are performed to determine the applicability of using data mining techniques on the many data sets to improve both process and product quality.