SLR Driven Heart Failure Prediction

Heart failure is one of the serious worldwide health issue which requires early identification and treatment. The paper offers an ensemble machine learning model that incorporates various patient variables, such as demographics, medical history, clinical measurements, and laboratory findings, to predict heart failure risk accurately. The model improves prediction accuracy and rectifies class imbalances by combining decision trees, random forests, and gradient boosting with SMOTE techniques. It also gives interpretable insights through feature significance analysis, allowing medical professionals to identify high-risk patients for immediate treatments and specialized therapy. Validated on heart failure data, the proposed model SLR (SVM +Logistic Regression + Random Forest Classifier) demonstrates significant predictive metrics, indicating its potential to enhance heart failure risk prediction and contribute to better clinical decision-making. The proposed methodology achieves 91.7% accuracy for the prediction of heart failure as it aimed to combine the best machine learning models which suites this dataset. This SLR model achieves precision, recall, F1 score with 0.93, 0.82 and 0.93 respectively which exhibits the efficacy of this model.