Dimensionality Reduction in the Creation of Classifiers and the Effects of Correlation, Cluster Overlap, and Modelling Assumptions.

The Atrium, University of Guelph Institutional Repository

Dimensionality Reduction in the Creation of Classifiers and the Effects of Correlation, Cluster Overlap, and Modelling Assumptions.

Show full item record

Title: Dimensionality Reduction in the Creation of Classifiers and the Effects of Correlation, Cluster Overlap, and Modelling Assumptions.
Author: Petrcich, William
Department: Department of Mathematics and Statistics
Program: Mathematics and Statistics
Advisor: McNicholas, Dr. Paul
Abstract: Discriminant analysis and random forests are used to create models for classification. The number of variables to be tested for inclusion in a model can be large. The goal of this work was to create an efficient and effective selection program. The first method used was based on the work of others. The resulting models were underperforming, so another approach was adopted. Models were built by adding the variable that maximized new-model accuracy. The two programs were used to generate discriminant-analysis and random forest models for three data sets. An existing software package was also used. The second program outperformed the alternatives. For the small number of runs produced in this study, it outperformed the method that inspired this work. The data sets were studied to identify determinants of performance. No definite conclusions were reached, but the results suggest topics for future study.
URI: http://hdl.handle.net/10214/2933
Date: 2011-08-31


Files in this item

Files Size Format View Description
William_Petrcich_THESIS.pdf 429.2Kb PDF View/Open MSc thesis

This item appears in the following Collection(s)

Show full item record

Search the Atrium


Advanced Search

Browse

My Account