Classification trees (CT) are nonparametric statistical methods. This method is a trees algorithm, performed to
diagnose the disease with the use of one or more risk factors. Although, in recent years availability of this method
especially for medical researches has increased, it is not common for the investigation of numerous risk factors of the
disease.
The aim of the study is to summarize theoretical knowledge related to CT and to investigate Restless Legs Syndrome
(RLS) risk factors with different approach. For this purpose, 206 patients, with whom a questionnaire form was
practiced in the University of Mersin, Faculty of Medicine, Department of Neurology, were included in this study and,
CT analysis is used for determination of variables whether these patients have RLS or not.
According to the analysis results, the variables which determine RLS disease was parallel with many risk factors
reported in literature.
Key Words: Classification and regression trees (CART), Diagnose, Decision trees, Misclassification error, Restless
Legs Syndrome.
|
