Diagnostic Classification Using Hepatitis C Tests

Abstract

This paper describes methods for automated classification of individuals by Hepatitis C medical category using data from a series of commonly used diagnostic tests. The methods are evaluated experimentally using a publicly available dataset from prior work. The accuracy of some methods compares favorably with similar results reported in prior work. In addition to quantitative results on prediction accuracy, training and testing times, and model sizes, the paper includes a detailed look at some concrete representative classifiers generated by a few of the competitive methods, permitting a human domain expert to further study the models and classifiers.