An application on Pharmacy Provision System data of Social Security Institution by logistic regression and CART analysis technics

One of the most widely used method of data-mining technics is classification and regression models. In this study, it was aimed at comparing the classification features of logistic regression, with Classification and Regression Trees (CART) algorithm which is one of the most widely used decision tree algorithm in data mining methods. In this context, an application are made with the aim of determining patient profile using penicillin group of antibiotics in antibiotics filling prescription for respiratory system diseases with the years of 2007?2009 data ensuring from pharmacy provision system of Social Security Institution. For the data set included, CART analysis was found to have a beter correct classification ratio than the logistic regression analysis.