Web Analytics Support System for Prediction of Heart Disease Using Naive Bayes Weighted Approach (NBwa)

Abstract

Datamining is a technique that uses several methods to find patterns or getting required information from database which can be used in decision support and predictions areas. In this research work an intelligent and effective system using Naïve Bayes modeling technique is analyzed for prediction of heart disease. User need to provide required values to the attributes for the application that is implemented as web based. The data is taken from database and relates trained data with user input value. Traditional systems cannot accurately find heart disease but this work can assist doctors to take correct decisions. Naïve Bayes for classification purpose to detect heart disease is used and this approach classifies output data as no, low, average, high and very high. Hence two basic functions namely classification and prediction is be performed. Accuracy of the system depends on algorithm and database used and with the Naïve Bayes Weighted Approach (NBwa) an accuracy of 86% is obtained.