Predicting the Coronary Heart Disease Risk Over the Next Decade through Ensemble Methodologies

Ascertaining diagnosis for Coronary Heart Disease (CHD) is a difficult procedure & must be completed precisely within a stipulated period. Numerous researchers have taken a gander at coronary illness risk factors & recommended specific Machine Learning (ML) & Artificial Intelligence (AI) models. This paper attempts to demonstrate the development of various predictive models for the risk of CHD over the next decade using the Framingham dataset with relevant features. The data undergoes thorough preprocessing, including handling missing values, outliers, & duplicates based on which a predictive model was built & evaluated. The work encapsulates essential steps, emphasizing the approach to predicting CHD & providing insights for potential healthcare applications. Attempting to improvise the predictive accuracy the work proposes implementing several ensemble approaches, combining the strengths of multiple models which showcased exceptional performance. Through the ensemble techniques, the individual models were integrated to mitigate weaknesses & enhance overall predictive performance that contributes to a more advanced and accurate prediction of CHD. The various strategic ensemble approaches enhance the reliability of the predictive model, offering a more effective tool for healthcare practitioners in assessing the likelihood of CHD in subjects.