Predicting Heart Disease Risk with Lifestyle Factors: Uncovering Vital Predictors via Feature Selection and Machine Learning Techniques

High incidence and mortality rates of cardiovascular diseases elevated them to the position of a major global public health challenge.Machine learning in predictive modeling became imperative for identifying those at risk of cardiac disease based on a host of factors, given the recent emergence of data-driven methodologies.It focuses on the role that lifestyle variables can play in predicting heart disease risk.The authors, in this paper, present a stacking ensemble method combining the strengths of various machine learning models to come up with a predictive model that would prove resilient and accurate.The methodology utilizes a single bagging or boosting ensemble model as the meta-learner and numerous traditional models as the base-learner.Furthermore, we implement a voting feature selection method to identify influential lifestyle predictors for heart disease by aggregating the results of five distinct techniques.By conducting a comprehensive comparative analysis of other machine learning models with and without reduced feature sets, the proposed ML technique's strengths in the context of heart disease prediction can be revealed.The results support the hypothesis that key lifestyle elements are important predictors of heart disease risk and that the proposed machine learning can effectively apply those features to health outcome predictions with an accuracy of up to 93.7%.This research contributes major knowledge for healthcare professionals and patients alike by putting emphasis on