Enhancing Cardiovascular Risk Prediction with Machine Learning Model

Heart failure (HF), a potentially fatal condition affecting over 26 million people globally, poses challenges for accurate prediction due to its complexity. This study compares the ability of two machine learning (ML) algorithms to predict Random Forest and Logistic Regression, utilizing extensive datasets containing health indicators and patient profiles. Key features include age, blood pressure, cholesterol, medical history, and various demographic, clinical, and physiological factors. Rigorous data preparation, encompassing standardization and handling missing values, ensures data reliability and model robustness.Assessment parameters such as accuracy, recall, and precision, are employed to evaluate the predictive capabilities of both models. Results indicate high accuracy for both models, with Logistic Regression (LR) at 88.14% and Random Forest (RF) at 98.91%. Random Forest outperforms Logistic Regression in precision and recall, showcasing its superior ability to discern positive and negative heart failure cases. This comparative analysis underscores the strengths of each model: Random Forest’s superior accuracy and robustness, and its simplicity in binary classification tasks. Healthcare professionals can leverage this knowledge to choose a model aligned with their specific goals, enhancing early detection, personalized therapies, and refined risk assessment in heart failure prediction. This study significantly advances the application of predictive modelling in healthcare.