Data mining cardiovascular risk factors

Nowadays, Medical centers collect various data in different diseases. Investigating these data and obtaining useful results and patterns with respect to the diseases are the aims of using these data. Great amount of these data and confusions results from, are problems to reach considerable conclusions. In this paper data mining is used to overcome this problem and to gain useful relationships among risk factors in cardiovascular diseases. With respect to the spreading and the share that these diseases have in death they have great importance. The technique used in this work is classification with decision trees and the software is used, is CART. The target variable (class) is LDL (Low Density Lipoprotein) and 27 predictor variables are used for the classification. By applying data mining on data about 1800 people in city of Isfahan, Iran results show that the most important variables with regards to LDL are of the level of cholesterol, age, body mass index, APOB, triglyceride, (APOB/APOA) and smoking.