CARDIOVASCULAR DISEASE PREDICTION SYSTEM USING MACHINE LEARNING.

ABSTRACT

Heart disease is ranked as one of the leading causes of death worldwide. Cases of cardiac diseases such as heart attacks and blood vessel disease are developing at an alarming rate. Therefore, it is critical to foresee any such ailments in advance. However, this is a challenging task that should be done accurately and quickly. Over the years, machine learning (ML) has helped to diagnose various diseases and classify or forecast their outcomes. It is widely known and has been acknowledged to play a significant role in the health sector. This paper employed machine learning algorithms and the Python programming language to predict the possibility of someone having heart disease. Instead of waiting to diagnose patients with heart disease, this system takes a more intuitive approach to predict its likelyhood before it occurs so as to aid early treatment, prevention, and a change of lifestyle where necessary. Its main objective is to help people know how susceptible they are to heart disease. It used survey data of over 300,000 U.S. citizens collected by the Center for Disease Control (CDC) to build a predictive model. Each feature contains the history of people's lifestyles over the period of one year, like BMI, smoking, etc. Two machine learning algorithms, namely Logistic Regression and Random Forests, and Support Vector Machines, were evaluated in this study, of which Logistic Regression had the highest performance of 76% accuracy with an AUC score of 0.84. In this research, a web-based cardiovascular system was developed which could serve as an efficient tool for cardiologist as well as non-cardiologist, aiming to provide an accurate result as the cases of cardiovascular disesase rises.