A Comparative Study on Heart Disease Prediction Using Data Mining Techniques and Feature Selection

Abstract

The world health organization shows us that cardiovascular disease is one of the noteworthy reasons for death in the world. In this paper, data mining classification techniques i.e. Naive Bayes (NB), Support Vector Machine (SVM), k-nearest neighbors' (k-NN), Decision Tree (DT), Neural Network (NN), Logistic Regression (LR), Random Forest (RF), Gradient Boosting are proposed to predict the probability of the coronary heart disease. In the present world, researchers are trying heart and soul to make advancements in the smart health care system. An automated system predicting the risk of heart disease may be added as a great achievement. This work of predicting heart disease is evaluated using the dataset from the UCI machine learning repository. The feature selection method enhances the performance of traditional machine learning algorithms. Among the classification algorithms, Random Forest (RF) algorithm with PCA has given the best accuracy of 92.85% for heart disease classification.