The Prediction of Heart Disease u sing Machine Learning Technique

Abstract

Medical errors are generally costly and harmful. There are many deaths worldwide every year. Clinical decision support systems provide opportunities to reduce medical errors and improve patient safety. Certainly, one of the most important aspects of applying such a system is the diagnosis and treatment of heart disease. Machine learning technology is implemented to analyze different kinds of heart-based problems. For this, this study essentially had two primary goals. Firstly, this paper intends to understand the role of variables in heart disease modeling better. Secondly, the study seeks to evaluate the predictive performance of the decision trees. Based on these results, first, men seem to be more susceptible to heart disease than women. Age, chest pain type, resting blood pressure, serum cholesterol, fasting blood sugar, maximum heart rate achieved, exercise-induced angina, ST depression induced by exercise relative to rest, the slope of the peak exercise ST segment, and the number of major vessels also show increased odds of having heart disease. Second, for the full model, the accuracy rate is 0.873, which implies that the error rate is 0.127. Among the patients who predicted not to have heart disease, the accuracy that would not have heart disease was 85.43%, and the accuracy that had heart disease was 89.10% among the patients predicted to have heart disease.