Integrated Mining for Cancer Incidence Factors from Healthcare Data

摘要

This paper describes how data mining is being used to identify primary factors of cancer incidences and living habits of cancer patients from a set of health and living habit questionnaires. Decision tree, radial basis function and back propagation neural network have been employed in this case study. Decision tree classification uncovers the primary factors of cancer patients from rules. Radial basis function method has advantages in comparing the living habits between a group of cancer patients and a group of healthy people. Back propagation neural network contributes to elicit the important factors of cancer incidences. This case study provides a useful data mining template for characteristics identification in healthcare and other areas.