Heart Disease Prediction and GUI Interaction based on Machine Learning

Abstract

Machine learning is now being used to detect heart disease. Considering that failure to diagnose a heart disease patient can lead to serious consequences, including delayed treatment, worsening of the condition, and even life-threatening situations, it is crucial to ensure that as many true patients as possible are confirmed. Thus, it is important to consider recall rates while maintaining a focus on accuracy. This article compares the application of four machine learning models, including Decision Trees (DT), Random Forests (RF), K-Nearest Neighbors (KNN), and Support Vector Machines (SVM), in heart disease prediction, and measures the effectiveness of these models by using accuracy, recall rates, and F1 score. The outcomes of the experiment reveal that the SVM model performs the best with a recall rate of 0.97. The balance of the model ensures that it achieves high recall without affecting accuracy. In addition, the author combines it with a Graphical User Interface (GUI) to achieve interactive effects. The model and its interactive functions selected in this experiment can easily avoid missing patients in the first screening and improve the accuracy of disease diagnosis.