Hypertension Multi-Year Prediction and Risk Factors Analysis Using Decision Tree

Hypertension is a widespread chronic condition affecting millions worldwide and is a leading cause of premature death. Early detection and prediction of hypertension are crucial for effectively managing and preventing hypertension-related complications. This study utilizes a multi-year dataset to develop a model for the early detection and prediction of hypertension using decision tree analysis. The study performed three experiments to identify the most significant risk factors for hypertension and analyze their impact on the condition’s development over multiple years. The first experiment revealed that Smart pruning, Deterministic ordering, and Deterministic sampling methods produced the best performance in terms of accuracy, precision, recall, F-measure, and Phi-coefficient. In the second experiment, the decision tree model’s performance was similar to other machine learning models in terms of accuracy and standard deviation but with much shorter training and scoring times. The third experiment identified age, systole, and BMI as the most significant risk factors associated with hypertension. These findings have the potential to aid in the development of effective prevention and treatment strategies for hypertension.