Cardiovascular disease is one of the most dangerous diseases that lead to death. It results from the lack of early detection of heart patients. Many researchers analyzed the risk factors of cardiovascular disease and proposed machine learning models for early detection of heart patients. However, these models suffer from high dimensionality of data and need to be improved in order to obtain highly accurate results. In this paper, we propose an operational proposed that can predict if the patient has cardiovascular disease or not. We test our proposed using five different standard datasets from the UCI repository. Our proposal consists of two main processes, the first process is the data preprocessing process, and the second is the prediction process. In data preprocessing, we prepare data for the prediction process, and moreover, we apply three different feature selection methods (e.g., PCA) to select the most relevant features from data. In the prediction process, we apply fourteen different prediction techniques (e.g., RF and SVM) over-employed datasets. We evaluate the employed techniques using four evaluation metrics: accuracy, precision, recall, and F1-score. The experimental results show that the LASSO method as a feature selection method with RF as a prediction technique produced the highest accuracy.