Survival Analysis of Heart Failure Patients with Advanced Machine Learning Models

According to world health organization reports, it is estimated that the deaths due to cardiovascular diseases (CVD) are rising every year across the world and have reached 1.79 crore ever. In many cases, cardiovascular diseases are caused due to ill healthy practices in daily routine such as smoking, physical inactivity, and consumption of alcohol in more frequent attempts, etc. Detection of CVD at the early stage is of significant requirement so as to reduce the risk of life and benefit the patient to undergo preventive medication and counseling. The work proposed in this paper is to adopt advanced machine learning models like Random Forest, AdaBoost, and XGBoost on a given data of heart failure patients who continued the follow-up at regular intervals without being affected by external constraints. In addition to the above-mentioned machine-learning methods, statistical survival analysis techniques like Kaplan–Meier and Cox regression are also used to identify the potential parameters that significantly affect the death rate.