A COMPARISON AND ANALYSIS OF SUPERVISED MACHINE LEARNING ALGORITHMS TOWARDS ACCURATE PREDICTING OF HEART DISEASES

Abstract

In machine learning, forecasting is one of the most significant applications. Machinelearning uses various techniques for prediction. Taking into consideration the recent work, we arefocusing on machine learning algorithms and analysing how these algorithms are used in thehealthcare industry to predict heart diseases. In supervised machine learning, the machine first statesthe patterns from labelled dataset (training dataset) and secondly it applies that on the unlabelleddataset (testing dataset) to predict the result. The training dataset includes input and correct output.Classification and Regression are the two techniques used in supervised machine learning.Classification technique is very commonly used to predict diseases in healthcare. Classification is alearning technique in which, deciding of class label to a given data done through machine learningalgorithms. Regression technique shows the relationship between two or more variables. We discoverlinks between dependent and independent variables through regression techniques.The core focus of this exploration is to conduct a systematic proportional study and examine of fourmachine learning algorithms, specifically random forest, support vector machine, KNN and decisiontree in heart disease prediction. We found that support vector machine and random forest provides thehighest accuracy in the prediction of heart disease among all. Random forest can be integrated withanother classifier to achieve more efficiency.