Evaluation of a Machine Learning Tool in the Domain of Genetic Host Reponses
Joel Pitt
Department of Computer Science
University of Canterbury
Abstract
This report examines the use of Support Vector Machines and Genetic Programming Classifiers (GPCs) to distinguish between classes of cancer based on gene expression data. The effect of feature selection on classifier accuracy and on the convergence time of GPCs is experimentally investigated, with the goal of making classi cation problems on gene expression data tractable to GPCs.