A Machine Learning Based Framework for Heart Disease Detection

Abstract

Even in rural parts of many nations, coronary heart disease has emerged as the main cause of mortality. More than 23 million people will die from cardiovascular disease by 2030, according to the World Health Organization (WHO). With the use of cardiovascular disease prediction, healthcare practitioners may check the characteristics necessary for diagnosis, such as blood pressure and diabetes, which are vital. Many data mining methods are currently used in the medical industry, but additional research is needed to evaluate how well these categorization approaches function in real-world settings. The project's purpose is to quickly identify the best candidates for constructing heart disease prediction models, and to do it in a timely manner. The goal of this study is to increase the accuracy of cardiac disease prediction by addressing and overcoming the issues in the area (CVDs). CAD systems, which help physicians make choices, are often developed as a result of breakthroughs in machine learning technology. The categorization and prediction of cardiac disease are discussed in this article. The algorithms explored include ANN, KNN, and CNN. To conduct the evaluation, we utilised the UCI Cleveland database. For these algorithms, a thorough evaluation of the utility and consistency of data mining approaches found that CNNs performed best. Usefulness and consistency of data mining techniques for these algorithms revealed that the CNN was the most reliable process.