Machine Learning-Based Prediction System for Risk Assessment of Hypertension Using Symptoms Investigations

Abstract

Hypertension is a common condition of cardiovascular disease that poses significant health challenges among the public on a larger scale globally. It is important to accurately predict the risk of hypertension to save people and improve overall quality of life. Traditionally, the detection of hypertension relies on clinical criteria such as blood pressure measurement and examination of medical history. However, these methods have drawbacks involving potential human error, time consumption, and the possibility of missed diagnoses. The paper aims to identify the features or symptoms of hypertension disease and predict its risk factors using machine learning algorithms. Apart from this, it is of utmost importance to identify the symptoms as they play a pivotal role in recognizing the type of risk for hypertension. To successfully conduct the work, a dataset of 13 attributes, including gender, age, smoking habits, etc, has been used, which is further visualized graphically to understand the pattern among them. Later, multiple machine learning-based learning techniques have been applied and examined on the basis of standard metrics. Results indicate that random forest models outperform existing approaches, achieving an accuracy of 87.26% in predicting low and high-risk hypertension. Furthermore, classification reports reveal superior precision, recall, and F1-score for random forests compared to alternative models. Insights from learning curves and confusion matrices provide a valuable understanding of model performance and data sufficiency. Overall, this research highlights the impact of machine learning in accurately predicting the risk of hypertension and underscores the importance of ongoing research efforts to translate these findings into practical clinical applications.