As per WHO around 20 million people die around the world due to cardiovascular disease. If proper medical care is provided, then many lives can be saved. The cause of heart attack in the health sciences sector is a difficult task. Large number of health care data is available for analysis. Different data mining algorithms are successfully applied to predict heart disease. Some of the data mining and machine learning techniques are used to predict the heart disease, such as Multilayer perceptron Neural Network (MLPNN), Decision tree, Naïve Bayes. This research paper compares three classification approaches, naive bayes, J48 and MLPNN, to predict heart disease from the give data set and studies the effectiveness of each method. Out if the three methods studied, MLPNN resulted with maximum correctly classified instances of data with a minimum root mean square error value, 0.14.
Heart disease prediction, MLPNN, J48, naive bayes, classification, data mining
