On Early Prediction and Diagnosis of Heart Disease Based on Random Forest Algorithm

This study addresses the early prediction and diagnosis of heart disease using the random forest algorithm to handle many features and perform classification. The study utilized CDC annual survey data related to the health status of 400,000 adults in the United States in 2020, including features such as gender, age, BMI, smoking status, alcohol consumption, history of stroke, and sleep duration. Based on this data, a random forest model was constructed and trained by dividing the data into training and testing sets. Following the model training, performance evaluation was conducted, including the calculation of accuracy and precision metrics. The experimental results demonstrated the excellent accuracy of our random forest model in predicting heart disease. This indicates that it can assist doctors in rapidly and accurately diagnosing and predicting heart disease, thereby improving patient treatment outcomes and quality of life, while also offering new possibilities for heart disease prediction.