Evaluation of Machine Learning Algorithms for Heart Disease Prediction in Healthcare
Shikha Prasher, Leema Nelson, Shanmugasundaram Hariharan
Abstract
Cardiovascular disease is most significant of the leading causes of mortality in the modern society. An important concern in clinical data processing is the diagnosis of heart disease. The enormous amount of information collected by the healthcare field has been established as helpful for using machine learning (ML) to help with judgement and prediction. In this work, four supervised machine learning techniques such as K-nearest neighbour (KNN), Multi-layer perceptron Decision Tree (DT) and REPTree have been used for predicting heart disease. The developed classifier model is tested with “Heart disease Prediction” obtained from Kaggle Repository. The performance metrics used to evaluate the classifier models are accuracy, precision, recall and f1-score. The existing four machine learning classifiers has been evaluated to forecast cardiac diseases and obtained maximum accuracy of 99.7% for KNN classifier.