This article proposes the use of three machine learning methods: Random Forest, Logistic Regression, and Neural Networks to predict cardiovascular diseases in order to provide recommendations and advice on the effective prevention and treatment of cardiovascular diseases. The models will be trained to learn the relationships between clinical factors and disease risk, and then evaluate their performance using a test set. Through experiments on the cardiovascular dataset at Tien Giang Provincial General Hospital, the results are presented, including performance evaluations of the models, comparing them to select the optimal method. The advantages and disadvantages of each method will be discussed, along with proposed improvements. The final model can be practically applied to support the diagnosis and treatment of cardiovascular diseases.