A Predictive Model for Heart Disease Detection Using Data Mining Techniques

Abstract

In this paper, the model is proposed to predict the heart disease detection by using data mining techniques. The data mining algorithm uses the Logistic Regression model and Neural Network model. The dataset of this paper uses the heart disease data at the University of California Irvine (UCI). There are a total of 303 Instances and 75 Attributes in the United States. The evaluation criteria using the confusion matrix table such as accuracy, precision, recall and F-Measure. The results show that the Logistic Regression model is better performance than Neural Network model. The Logistic Regression model has 95.45% precision and 91.65% accuracy. The web application can be support for the user, who wants to diagnose heart disease detection.