The PREDICTION FOR HEART DISEASE USING DIVERSE MACHINE LEARNING APPROACHES AND TECHNIQUES

Abstract

The heart is the most important organ of the human body. There are two main functions of the heart firstly, to collect blood from tissues of the body and pump it to the lungs,and second, to collect blood from the lungs and pump it to all tissues of the body¹.Many people have died because of heart disease. Therefore, it is important to predict that disease at the right time. By using machine learning and data mining techniques diseases can easily be predicted and diagnosed. Wearable sensor devices also can be used in the Internet of Things, and streaming systems². The main objective of this research is to analyze core machine learning algorithms for heart disease prediction,for instance, SVM (Support Vector Machine), and Logistic Regression.K-Nearest Neighbors Algorithm, Decision, and Random Forest.Our Trained model for Logistic Regression showed 83% accuracy prediction result whereas the Decision Treealgorithm showed only70%which is 13% less than Logistic Regression.The result of the K-Nearest Neighbors Algorithmis 84% whereas SVM showed90% accuracy prediction result which is quitebetter than previously used algorithms. Then Random Forest showed 91% result which is a better result than all previously used algorithms i., eDT (Decision Tree), RF (Random Forest) and K-Nearest Neighbors Algorithm, Python Programming jupyterNotebook which is excellent in code and data.Hlaudi Daniel Masetheet al.