Cardiovascular Disease Prediction using Machine Learning Methods

Abstract: Heart is the most important organ for all living organisms. The heart related diseases caused numerous numbers of deaths worldwide from past few decades. Prediction and diagnosis of the heart diseases with very high precision and correctness is required for early diagnosis. Machine Learning helps in making predictions and decisions from large data sets of data consisting medical parameters. The paper demonstrated many Machine Learning algorithms such as Decision Tree Classifier, KNeighbors Classifier, Naive bayes, Random Forest Classifier, Grid Search CV, SVM for predicting the heart disease using Erbil heart disease dataset from kaggle having 22 different medical and non-medical attributes. The precision, accuracy, F1-score, and recall of all algorithms used for predicting the heart disease is evaluated. Decision Tree Classifier algorithm provided a good accuracy of 98% among all other algorithms.

Keywords: Machine learning, Naive bayes, KNeighbors Classifier, Random Forest Classifier, Grid Search CV, SVM, Decision Tree Classifier, F1-score, Confusion matrix.