Enhanced Cardiovascular Disease Prediction Using Advanced Machine Learning with Hybrid Feature Selection

Cardiovascular diseases (CVDs) are the leading cause of mortality globally, necessitating advanced predictive models for early detection and prevention. This paper introduces a machine learning-based approach for predicting CVDs, leveraging a hybrid feature selection method to enhance model accuracy and interpretability. The methodology includes preprocessing patient data through normalization, one-hot encoding, and advanced imputation, followed by splitting the dataset into 80% training and 20% testing sets. A hybrid feature selection approach starts with a filter method using mutual information to reduce dimensionality, followed by a wrapper method with Random Forest to refine feature sets through cross-validation. The model integrates three machine learning algorithms: Random Forest, Support Vector Machine (SVM), and Neural Networks, each fine-tuned for optimal performance. Validation is conducted using 10-fold cross-validation within the training set, and the model is evaluated on the testing set using metrics such as accuracy, precision, recall, F1-score, and the area under the ROC curve. Results demonstrate that the proposed model significantly outperforms traditional methods, with the Random Forest achieving an accuracy of 85%, SVM 87%, and Neural Networks 89%. The advanced feature selection method contributed to the model's efficiency and effectiveness, leading to superior performance metrics. This study highlights the potential of integrating advanced machine learning techniques with hybrid feature selection in medical diagnostics, offering a robust tool for early and accurate CVD prediction.