Cardiac disease is a critical health illness that has the potential to cause fatal complications such as heart attacks. Early diagnosis, intervention, and management are essential for the treatment of this condition. This study as to find the feasibility of utilizing data-mining approaches in big data to obtain clinically actionable information for diagnosing heart disease. However, more emphasis is placed on feature selection to enhance the accuracy of the prediction by reducing the number of crucial features. Some of these algorithms like random forest, logistic regression, support vector machines (SVM), and multilayer perceptron is for heart disease prediction. Other processes, such as bagging and boosting, belong to the category of ensemble methods that help to increase the accuracy of the models. In addition, a cross- and ensemble approach that combines Random Forest with other models, including logistic regression and transfer learning models, is discussed. The performance of the model was evaluated using performance metrics, such as accuracy, precision, recall, and F1 score. Among the models analyzed, some offered up to 95% accuracy in predicting heart disease. Classifiers used together with several machine-learning algorithms have been shown to enhance the accuracy of predicting heart disease.