Cardiovascular Disease Prediction

Abstract

An illness that negatively impacts the heart and blood vessels is referred to as having a cardiovascular disease. Due to the fact that it is one of the leading sources of mortality worldwide, early prediction is required. Prediction and classification issues are frequently addressed using machine learning. Therefore, we attempted to create a system that can identify cardiovascular disease in its early stages so the person can be informed beforehand, which will aid in an early diagnosis. We had 12 features , which included the age, gender, the type of chest pain, resting blood press. , cholest., maximum heart rate, resting Bp, fasting blood sugar, exercise angina, ST slope and old peak, we have taken a dataset from the IEEE Data Port. We then reduced the features, after finding the correlation between them, and employed six machine learning methods, including SVM, decision tree, random forests, KN-neighbour, XG Boost and multilayer perceptron and also combined certain models together then evaluated the model on basis of it’s accuracy, sensitivity , precision, F1 score , log loss and Mathew’s correlation coefficient . We concluded that the random forest and MLP Model gave the highest accuracy of 91%