Date of Award

Summer 2009

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Computational Modeling & Simulation Engineering

Program/Concentration

Modeling & Simulation Engineering

Committee Director

Jiang Li

Committee Member

Frederic McKenzie

Committee Member

Yuzhong Shen

Call Number for Print

Special Collections LD4331.E58 J53 2009

Abstract

With the current development of proteomics techniques, the discovery of potential molecular biomarkers for early detection of prostate cancer has been greatly improved. In this thesis, we implemented five classifiers including the support vector machine (SVM), Bayesian Classifier, Decision Tree, Random Forest and the Multilayer perceptron (MLP) to test their effectiveness in prostate cancer biomarker identification using matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) tissue imaging data. The classifiers were utilized to discriminate cancer mass spectra from normal ones for the data collected at the Eastern Virginia Medical School (EVMS). There are 94 7 spectra from normal tissues and 27 from cancer tissues (total 974 samples). The pipeline of biomarker identification consists of a series of processing steps including classification using the classifiers. This thesis focused on comparing these five classifiers for the classification task. We also tested several committee machines for the classification by combining different classifiers. Experimental results showed that the Bayesian classifier outperformed the other four classification algorithms and committee machines can provide even better performances.

Rights

In Copyright. URI: http://rightsstatements.org/vocab/InC/1.0/ This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).

DOI

10.25777/xdsr-gh84

Download

Share

COinS