Unveiling Heart Disease Using Data Mining and ML Models

Computer Science and Engieering

—Heart disease remains a significant global health concern, highlighting the critical need for accurate predictive models to enable timely interventions and improve patient outcomes. This study delves into the realm of machine learning and deep learning techniques for predicting heart disease using a dataset encompassing various clinical parameters. Initially, the study employs feature selection methods such as SelectKBest, LassoCV, and correlation analysis to pinpoint pertinent features. Subsequently, a range of classification algorithms—including K-Nearest Neighbors (KNN), Random Forest, AdaBoost with Random Forest, Gradient Boosting, XGBoost, as well as deep learning models like Dense Neural Networks (DNN) and Long Short-Term Memory (LSTM) networks—are trained and assessed. Through hyperparameter tuning and cross-validation strategies, model performance metrics such as accuracy, recall, precision, and F1 score are optimized. The experimental outcomes highlight the efficacy of the proposed models, with the top-performing model achieving an accuracy of 97.82%, precision of 98%, recall of 1, and F1 score of 0.98. Additionally, leveraging deep learning models for feature extraction yields promising results when integrated with traditional machine learning algorithms. This study contributes significantly to advancing heart disease prediction methodologies and underscores the potential impact of machine learning and deep learning in healthcare analytics. Outcome Assessment —Machine learning models, notably the K-Nearest Neighbors (KNN) algorithm, play a crucial role in improving healthcare outcomes, particularly in the early detection of heart disease, which has a significant impact on patient survival rates. This study highlights KNN as the most effective model, achieving an impressive accuracy of 97.82%, precision of 98%, recall of 100%, and F1 score of 98%. These results outperform existing methods, underscoring the KNN algorithm's effectiveness in predicting heart disease. Utilizing a comprehensive dataset containing vital clinical parameters, the KNN model demonstrates robust performance, showcasing its potential for practical clinical applications. Furthermore, this research underscores the importance of precise feature selection and thorough model evaluation techniques in optimizing predictive accuracy, paving the way for enhanced healthcare analytics and patient care.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.