Heart disease prediction system using linear regression, smoreg, and rep trees algorithms

Abstract

The broad area of data mining covers most of the fields of research. Its role in medical diagnosis is very motivative to the researchers. It is very easy for the medical practitioners to analyze and treat the disease of the patients at an early stage. The proposed work deals with predicting heart disease of the patients at an early stage. The method was organized in three stages, Data collection, Data preprocessing and Data classification. The dataset for the work was collected from UCI repository. The collected sample was first preprocessed to clean unwanted information from the dataset. Classification operation is then performed on the preprocessed data. Classification is carried out with three different techniques, Linear regression model, SMOreg and REP trees. The results of the three methods were compared based on Root mean squared error and the Absolute error and are tabulated.