Enhancing Heart Disease Prediction Using Different Ensemble Models

Heart-Disease is one of the main causes of death globally, early detection and treatments require precise and effective prediction models. The capacity of traditional approaches to manage intricate datasets and produce accurate forecasts is frequently constrained. Using a variety of machine learning classifiers as well as ensemble classifiers, such as Logistic Regression, Random Forest, Decision Tree, K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Naive Bayes, Gradient Boosting and Multilayer Perceptron, this work aims to improve the prediction of heart attack and stroke. Important cardiovascular risk factors are represented by the dataset's properties, which include age, sex, cholesterol levels and more. Following feature engineering, preprocessing and exploratory data analysis (EDA) many models were trained and assessed. The maximum accuracy (99.17%) was attained by Multilayer Perceptron, Random Forest and Decision Tree achieved 78.68% accuracy, followed by SVM (81.96%), Gradient Boosting (73.77%), Naive Bayes (81.97<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup>), Logistic Regression (85.12%) and KNN (62.30%). The potential benefits of these models for real-world healthcare applications were demonstrated by the implementation of a continuous forecasting system that used streamlit UI to enable user interaction and deliver rapid forecasts.