On the Discovery of Patterns in Medical Data

Jorge C. G. Ramirez, Lynn L. Peterson, Dolores M. Peterson, Gretchen K. Cormier

We have looked at the KDD field for techniques that might be useful in our domain. As a result of our investigation to date, we have concluded first that KDD is pervaded by human intervention and verification of hypotheses, as opposed to discovery. Second, data mining techniques require that data be in standard form, whether it be in the form of a training set or preprocessed to meet the techniques’ input requirements. Finally, our perspective of spanning the course of disease does not well fit with any of the traditional data mining techniques. In view of these conclusions, it seems that the temporal issue is the dominant factor in our problem domain and therefore, we must attempt to adapt or create new techniques to accomplish our goal.

This page is copyrighted by AAAI. All rights reserved. Your use of this site constitutes acceptance of all of AAAI's terms and conditions and privacy policy.