An Adaptive Computer-Based System for the Prescription of Warfarin
Mark Mayo
Department of Computer Science
University of Canterbury
Abstract
This report presents the development of an algorithm for the prediction of the effects of Warfarin, and thereby its prescription. The use of Machine Learning techniques, includind Genetic Algorithms and Artificial Neural Networks, are investigated, and it is demonstrated that Machine Learning can be used to accurately predict changes in the Warfarin dosage requirements of individial patients, as well as to predict their INR levels when given a particular dosage, through the use of an adaptive patient model.