Classification Technique for Heart Disease Prediction in Data Mining

Currently, a lot of methods have been implemented for prediction analysis. In the proposed study, clustering and classification of the input information for heart disease forecasting is executed with the help of k-means clustering algorithm and SVM (support vector machine) classification model on the basis of prediction analysis methods. The back propagation algorithm along with k-means clustering algorithm is applied for the clustering of information. These algorithms support to enhance the precision of prediction analysis. A heart disease data suite obtained from the UCI repository is used for judging the performance of proposed algorithm. This data suite comprises total 76 features. But, whole tests require a subset of 14 features. Particularly, Cleveland database particularly is utilized by the machine learning researchers throughout the tests. A comparison between proposed study and earlier method (using arithmetic mean) is performed in terms of certain parameters such as accuracy, error recognition rate and execution time.