Machine Learning Based Diagnosis System for Easy and Fast Heart Disease Prediction

The goal of the present research is the designing of Machine learning (ML) based system for heart disease (HD) prediction. In the study, different ML algorithms such as K-nearest neighbor, Decision Tree, Support Vector Machine, Logistic Regression, Random Forest, and Naïve Bayes have been used. Besides, some feature selecting methods such as Gain Ratio, Correlation-based feature selection, OneR attribute were applied to select the most influential features from Cleveland HD dataset. In addition, an extensive preprocessing measure such as normalization, integration, and discretization were applied to increase the predictive accuracy of the study. The study outperforms the state-of-art research once classified with different preprocessing and feature selection methods.