Predicting the occurrence of essential hypertension using annual health records

Abstract

Essential or primary hypertension is one of the major diseases affecting health in most societies. Hypertension can be controlled if its propensity can be forecasted for an individual, allowing necessary changes to be made in life-style. In this paper, we have used the data of subjects who had no prior history of hypertension for 2 or more consecutive years and applied various classifiers to predict occurrence of essential hypertension. For experimentation, we have used medical check-up data of Korean population between 2002-2013 and achieved 80% accuracy.