Further Study on the Identification of Predictive Capability of Classifiers for Early Heart Disease Detection Using Machine Learning

Abstract

Amongst all fatal diseases, cardiovascular diseases are considered the most prevalent. Due to the increase in workload, unhealthy diets and fast paced lifestyles, the younger generations have also fallen victim to heart complications. However, the early diagnosis of cardiovascular diseases can help in making decisions on lifestyle changes in high-risk patients and reduce the complications associated with it. The proposed system addresses this issue by using data collected by the Healthcare industries around the world and are used to effectively make predictions. The results from prediction of the system are used to prevent the disease and thereby reduce the cost for surgical treatments and other expensive tests associated with it. This prediction system aims to aid such individuals in addition to issues such as lack of physicians in rural areas and places with low quality of healthcare. By providing a prediction model for heart diseases at an early stage, this system helps reduce the cost of medical tests and the errors associated with it are also considerably reduced compared to manual testing. The added feature of instant diagnosis can be very useful in case of an emergency. The accuracy of machine learning algorithms is checked and the capability of deep learning classifiers are checked for cardiovascular disease identification and prediction along with a rigorous process of data mining to remove noisy data for a better decision making system with an extremely effective accuracy.