A Study of Heart Disease Prediction Based on Logistic Regression and CART Classification

According to the World Health Organization, cardiovascular and cerebrovascular diseases are still the world's leading cause of death and disability. More than 500 million people worldwide are affected by cardiovascular and cerebrovascular diseases, resulting in a large number of deaths and disabilities every year. Heart disease is the main cause. The early symptoms of heart disease are often not obvious, which can easily lead to missed diagnosis and misdiagnosis. Therefore, it is crucial to improve the accuracy of the prediction of heart disease, thereby improving the accuracy and convenience of diagnosis. The purpose of this study was to use two machine learning algorithms, Logistic Regression and CART classification, to predict the risk of heart disease. By collecting clinical data of heart disease patients, including key indicators such as age, gender, blood pressure, cholesterol levels, etc., a predictive model was constructed. Logistic regression was used to assess the association between each factor and the incidence of heart disease, and to quantify its impact. The CART classification method generates a decision tree model by recursively segmenting the data, and intuitively displays the decision path for predicting heart disease. Comparing the prediction performance of the two models, the results show that both models can effectively identify individuals at high risk of heart disease, the CART model shows more flexibility in capturing complex nonlinear relationships, and the logistic regression model is more interpretable, the model is simpler, and the binary classification problem is also very effective. The optimal model selected in this paper is the logistic regression model, which not only provides a scientific basis for the early warning system of heart disease, but also lays a data foundation for further exploring the pathogenesis of heart disease and personalized treatment strategies.