Cardiovascular Disease Prediction and Comparative Analysis of Varied Classifier Techniques

Heart disease is one of the leading causes of death in the world. Early detection is a challenging task for medical practitioners to save lives and reduce the mortality rate of patients. This work studies the ten machine learning (ML) classification algorithms namely Random Forest (RF), K-Nearest Neighbor(K-NN), Decision Tree (DT), Gaussian Naïve Bayes (GNB), Support Vector Machine (SVM), Adaboost, Stochastic Gradient Decent (SGD), Logistic Regression (LR), Artificial Neural Network(ANN) and Linear Discriminant Analysis (LDA) on two different medical datasets such as heart disease UCI dataset with 14 subsets of attributes and Kaggle dataset with 16 attributes for prediction. For validating the prediction model k-fold cross-validation resampling method is applied. The varied metrics such as specificity, accuracy, sensitivity, false discovery rate (FDR), Recall, Net present value (NPV), Matthews Correlation Coefficient (MCC) and F-Measure are considered to evaluate the performance of the classification algorithms. Comparatively, LDA performed better over the other algorithms with 96.54% accuracy on the UCI dataset and 95.58% accuracy on the Kaggle dataset.