The Prediction of Hypertension Risk

Abstract

This article presents an estimation of the hypertension risk based on a dataset on 1007 individuals. The application of a Tobit Model shows that “Hypertension” is positively associated to “Age”, “BMI-Body Mass Index”, and “Heart Rate”. The data show that the element that has the greatest impact in determining inflation risk is “BMI-Body Mass Index”. An analysis was then carried out using the fuzzy c-Means algorithm optimized with the use of the Silhouette coefficient. The result shows that the optimal number of clusters is 9. A comparison was then made between eight different machine-learning algorithms for predicting the value of the Hypertension Risk. The best performing algorithm is the Gradient Boosted Trees Regression according to the analyzed dataset. The results show that there are 37 individuals who have a predicted hypertension value greater than 0.75, 35 individuals who have a predicted hypertension value between 0.5 and 0.75, while 227 individuals have a hypertension value between 0.0 and 0.5 units.

Fuzzy c-Means Clusterization and ANN- MLP Prediction of Malign Breast Cancer in a Cohort of Patients

“…Augmented data can be applied to predict main diseases [41]. Machine learning algorithms can also be used to predict hypertension risk [42] and diabetes [43]. Finally, telemedicine is an essential tool in the passage from industry 4.0 to industry 5.0 [44].…”

The Prediction of Diabetes

“…On a methodological point of view, the usage of augmented data can improve the application of predictive algorithm even in the management of healthcare datasets [72]. Similar considerations can also be realized for other disease that massively affect the global population such as hypertension [73]. Furthermore, the usage of machine learning can be effectively improve the ability to predict diabetes based on glycemic status of patients [74].…”