Application of the Decision Tree Algorithm for Early Detection of Heart Disease Based on IoT

Abstract

Heart disease is one of the leading causes of death worldwide, accounting for 32% of all global deaths. Technological developments, particularly in the Internet of Things (IoT), enable real-time monitoring of heart health and early warning alerts. This study aims to implement a Decision Tree algorithm to classify patient conditions based on vital parameters, including BPM, SpO₂, systolic and diastolic blood pressure, and body temperature. The model was trained using a vital parameter dataset and evaluated using a confusion matrix, ROC curve, and feature importance. Test results show that the Decision Tree model achieves an accuracy of 85% with a macro-AUC value of 0.448. These results prove that the Decision Tree algorithm can be used for patient condition classification with reasonably good performance, although the model still tends to make prediction errors in some minority classes.