Machine Learning Based Classifier for Risk Prediction of Cardiovascular Diseases with Progressiveness

Abstract

Heart disease is a significant medical issue and it influences countless individuals. newlineCardiovascular Disease (CVD) is one such critical health ailment. There is no newlinesignificant research that focus on powerful prediction tools to find relationships and newlinetrends in data in the medical division. CVD is such a significant health issue that is newlineaffecting the entire world, particularly in low as well as middle-income countries. newlineWith the pace it is growing, it will continue as the main cause of mortality in the newlineupcoming twenty years. To extract valuable data in case of heart disease prediction, newlinedifferent types of machine learning as well as deep learning methods have been newlineintroduced in the past. However, the accuracy of the past results is not acceptable. newlineUsing novel and robust machine learning and deep learning techniques, this research newlinesuggests a disease prediction system for CVD. newlineThe current advancement in Machine Learning and Deep Learning field has newlinemotivated researchers to adopt advanced methods for early prediction of heart-related newlinedisease. The entire thesis can be split into four phases: newlineThe first phase focuses on the development of an ensemble approach that gives newlineprecedent and direction for the evolution of a new type of risk prediction method in newlinethe CVD field. There are two contributions of this phase such as- 1) To get more newlineaccurate results with the latest heart disease dataset that utilizes recent machine newlinelearning approach. The machine learning technique is to predict the severity of newlinepatient s heart disease. 2) To collect heart disease records of patients from different newlinedata sources, then analyze them to predict the severity of diseases. In order to obtain newlinethe ensemble model, this research has suggested incorporating three different newlineclassifier modules namely Gaussian Naïve Bayes (GNB), Decision Tree (DT), and KNearest Neighbour (KNN) classifiers. The outcome of these modules is processed newlinethrough the decision-making process to obtain the final outcome. Ensemble of GNB, newlineDT, and KNN reported the overall performance