Abstract:
The digital revolution has brought an explosion of stored
data to our world, and data mining, especially on the
rapidly expanding medical databases, has the capabilities
to turn this information into new and useful medical
knowledge. Data mining with cost sensitive decision trees
created using Genetic Programming is focused upon, with
an emphasis on their relevence and potential in medical
data mining. An application that uses elitist multiobjective
Genetic Programming to obtain a pareto front of univariate
and linear decision trees was implemented, and various
mutation operators were tested and compared on their
performance in the development of the decision trees.
Acknowledgements